[Senate Hearing 110-1179]
[From the U.S. Government Publishing Office]
S. Hrg. 110-1179
PRIVACY IMPLICATIONS OF ONLINE ADVERTISING
=======================================================================
HEARING
before the
COMMITTEE ON COMMERCE,
SCIENCE, AND TRANSPORTATION
UNITED STATES SENATE
ONE HUNDRED TENTH CONGRESS
SECOND SESSION
__________
JULY 9, 2008
__________
Printed for the use of the Committee on Commerce, Science, and
Transportation
U.S. GOVERNMENT PRINTING OFFICE
76-329 WASHINGTON : 2012
-----------------------------------------------------------------------
For sale by the Superintendent of Documents, U.S. Government Printing Office,
http://bookstore.gpo.gov. For more information, contact the GPO Customer Contact Center, U.S. Government Printing Office. Phone 202�09512�091800, or 866�09512�091800 (toll-free). E-mail, [email protected].
SENATE COMMITTEE ON COMMERCE, SCIENCE, AND TRANSPORTATION
ONE HUNDRED TENTH CONGRESS
SECOND SESSION
DANIEL K. INOUYE, Hawaii, Chairman
JOHN D. ROCKEFELLER IV, West TED STEVENS, Alaska, Vice Chairman
Virginia JOHN McCAIN, Arizona
JOHN F. KERRY, Massachusetts KAY BAILEY HUTCHISON, Texas
BYRON L. DORGAN, North Dakota OLYMPIA J. SNOWE, Maine
BARBARA BOXER, California GORDON H. SMITH, Oregon
BILL NELSON, Florida JOHN ENSIGN, Nevada
MARIA CANTWELL, Washington JOHN E. SUNUNU, New Hampshire
FRANK R. LAUTENBERG, New Jersey JIM DeMINT, South Carolina
MARK PRYOR, Arkansas DAVID VITTER, Louisiana
THOMAS R. CARPER, Delaware JOHN THUNE, South Dakota
CLAIRE McCASKILL, Missouri ROGER F. WICKER, Mississippi
AMY KLOBUCHAR, Minnesota
Margaret L. Cummisky, Democratic Staff Director and Chief Counsel
Lila Harper Helms, Democratic Deputy Staff Director and Policy Director
Christine D. Kurth, Republican Staff Director and General Counsel
Paul Nagle, Republican Chief Counsel
C O N T E N T S
----------
Page
Hearing held on July 9, 2008..................................... 1
Statement of Senator Carper...................................... 80
Statement of Senator DeMint...................................... 71
Statement of Senator Dorgan...................................... 1
Prepared statement of Hon. Daniel K. Inouye.................. 3
Prepared statement by Hon. Ted Stevens....................... 68
Statement of Senator Klobuchar................................... 68
Statement of Senator Nelson...................................... 75
Statement of Senator Thune....................................... 72
Statement of Senator Vitter...................................... 66
Witnesses
Crews, Jr., Clyde Wayne, Vice President for Policy/Director of
Technology Studies, Competitive Enterprise Institute........... 47
Prepared statement........................................... 48
Dykes, Robert R., CEO, NebuAd, Inc............................... 17
Prepared statement........................................... 19
Harris, Leslie, President/CEO, Center for Democracy and
Technology..................................................... 22
Prepared statement........................................... 24
Hintze, Michael D., Associate General Counsel, Microsoft
Corporation.................................................... 55
Prepared statement........................................... 57
Horvath, Jane, Senior Privacy Counsel, Google, Inc............... 11
Prepared statement........................................... 12
Kelly, Chris, Chief Privacy Officer, Facebook, Inc............... 40
Prepared statement........................................... 42
Parnes, Lydia B., Director, Bureau of Consumer Protection,
Federal Trade Commission....................................... 4
Prepared statement........................................... 5
Appendix
Letter, dated July 9, 2008, to Hon. Daniel K. Inouye and Hon. Ted
Stevens from Hon. Richard Blumenthal, Attorney General, State
of Connecticut................................................. 87
Response to written questions submitted by Hon. Maria Cantwell
to:
Robert R. Dykes.............................................. 97
Leslie Harris................................................ 96
Jane Horvath................................................. 93
Chris Kelly.................................................. 92
Lydia B. Parnes.............................................. 87
Response to written questions submitted by Hon. David Vitter to:
Clyde Wayne Crews, Jr........................................ 107
Robert R. Dykes.............................................. 98
Leslie Harris................................................ 97
Michael D. Hintze............................................ 108
Jane Horvath................................................. 95
Chris Kelly.................................................. 93
Lydia B. Parnes.............................................. 90
PRIVACY IMPLICATIONS OF ONLINE ADVERTISING
----------
WEDNESDAY, JULY 9, 2008
U.S. Senate,
Committee on Commerce, Science, and Transportation,
Washington, DC.
The Committee met, pursuant to notice, at 10 a.m., in room
SR-253, Russell Senate Office Building, Hon. Byron L. Dorgan,
presiding.
OPENING STATEMENT OF HON. BYRON L. DORGAN,
U.S. SENATOR FROM NORTH DAKOTA
Senator Dorgan. We are going to begin the hearing this
morning. This is a hearing of the Senate Commerce Committee. I
am going to begin the hearing now. We will have other Senators
join us, but at the moment there is scheduled a series of five
votes beginning at 11:15. That may slip a bit. It may slip to
11:30, which means that I would have until 11:45 to leave the
room in order to still make the vote. That would give us an
hour and 45 minutes. If we do not finish the hearing in an hour
and 45 minutes, we will have to recess. The recess will last at
least an hour to an hour and a half because five votes take
that long.
So my hope would be that we can finish this hearing in
about an hour and 45 minutes. I do not want to shortchange the
subject. This is a very important subject. We will be holding
another hearing on this subject, but for now, as an opening, I
want to at least give an opening statement, and then I am going
to call on all of the witnesses to provide the testimony. And
then we will have time for questions by Senators who come to
the hearing.
I want to thank all of you for joining us today to discuss
an important topic of privacy in the context of marketing for
online advertising. The Commerce Committee has always had an
interest in this subject of protecting privacy and doing so in
a way that is thoughtful and appropriate. And we need to take a
closer look, I think, at Internet users' privacy as the field
of online advertising develops.
I understand, first of all, there are many, many benefits
to online advertising. I understand that the free Internet and
the open architecture of the Internet allows Internet service
providers and Internet companies to provide services and
products and information in a way that is almost breathtaking,
and I understand the backbone of much of that is supported by
advertising, by online Internet advertising.
The questions that we discuss and raise today are not meant
to suggest that advertising has no value as it relates to the
Internet. Quite the contrary is the case. But I think there are
issues that are developing that are important issues and those
are issues about the invisibility of data collection, the
collection of information about online users, the security of
the information that has been collected, and the use of that
information. I am concerned about the ability of users of the
Internet to choose to allow others access to their data with an
understanding of where it will be transferred and how it will
be used. I am concerned about the users' ability to control
this information.
Most of the discussion about advertising on the Internet
these days is about--not all of the discussion, but most of it
is about behavioral advertising. The companies that have been
gathering information about those who use the Internet are
companies that wish to find ways to better target advertising.
I was actually this morning visiting with my college-aged
daughter about this subject, and she was asking about the
hearing. And I said, well, think about going to a big shopping
center and going to four stores, and you stop at the cosmetics
section, you stop at the shoe section, you shop for a dress,
and you go to CVS, and there is someone behind you with a
notebook making notes about every place you are stopping and
the products you are searching and looking at. That becomes
part of a data bank they send to someone.
Well, that is what happens in many cases now with respect
to your use of the Internet. Someone is gathering information
about where you traveled and what you viewed, and that goes
into a data bank and it can be sold or resold and becomes the
process by which companies will use the information in which to
target advertising directly to you.
I think that the issue here of privacy, who is collecting
what kind of information, how does that information exist in
what identifiable form or how might that be used, who is it
sold to, does the consumer know, all of those issues I think
are very, very important. And the issues that surround them I
think need to be discussed by the Congress in the context of
trying to decide are there protections, further protections,
necessary, not only what protections now exist, but are further
protections necessary.
Companies believe that by gathering information about
online users, they can serve more relevant advertisements to
individuals and increase the amount that they can charge per
ad. And I understand all that. And revenue is important to the
operation of diverse websites, but it is also important for us
to ask the questions about protection for users' privacy and
whether they should have a choice about whether they want to
have data shared about what they are doing on the Internet.
There is some discussion about self-regulation, and the
principles proposed by the Federal Trade Commission I think
some would suggest are a good start. Some would suggest they
are short of what is necessary. The FTC might need to go
further and ensure the enforcement of any guidelines. Congress
may need to address our patchwork of privacy laws because all
of this is a developing area, and when you are talking about
the individual's right to privacy, it is very, very important.
I do want to mention additionally that I had invited the
Internet service providers to testify today, and they declined
the invitation. So I am going to do another hearing and that
hearing will only be with the Internet service providers
because I think they have to be a part of this discussion.
I do appreciate all of those who have decided to come at
our invitation and give us their perspective on these important
issues. And we are going to begin today by hearing from Lydia
Parnes, who is the Director of the Bureau of Consumer
Protection at the Federal Trade Commission. Then I have asked
the panelists to be seated at the table, and we are going to
hear from all of those on panel two and then we will ask
questions.
I say to Senator Vitter I do not want to do opening
statements. I will be happy to recognize you for a moment, but
what I want to do--we have a vote starting at 11:15. What I
would like to do--if that slips, we may be able to stay here
until 11:45, but I want to get all the witnesses to testify and
then have an opportunity for questions.
Senator Vitter. All right.
Senator Dorgan. All right.
Ms. Parnes, you have testified before this Committee
before. We appreciate your being here again. Let me ask
consent, as I recognize you, for the statement by Senator
Inouye to be made a part of the permanent record.
[The prepared statement of Senator Inouye follows:]
Prepared Statement of Hon. Daniel K. Inouye, U.S. Senator from Hawaii
In the United States of America, privacy is a treasured right, but
it is also a right that seems to come under regular attack.
Today, commercial entities using digital means can track nearly all
of our marketplace moves. Websites and Internet service providers can
watch where we go online, what we purchase over the web, and where we
linger on the Internet.
Too many consumers spend time on the Internet without knowledge or
notice that they are under commercial surveillance. They assume they
are in the privacy of their own home and that this privacy will be
respected. Unfortunately, this is not always the case.
I am troubled by the current state of affairs. I fear that our
existing patchwork of sector-specific privacy laws provides American
consumers with virtually no protection. At the same time consumers in
other countries are treated with more respect and concern by the very
same companies who so freely collect our most private information
without warning.
American consumers deserve better. With so much of our commerce and
entertainment migrating to the Internet, consumers should not be asked
to surrender their privacy each time they go online.
Ensuring that every consumer's right to privacy is appropriately
protected will require the Congress's continued attention. Today's
hearing on privacy and online advertising represents merely a start,
and I look forward to holding additional hearings on these matters
later this year.
Senator Dorgan. Ms. Parnes, you may proceed. We have asked
all of you to take 5 minutes for your oral testimony, and the
permanent record will include your full statements.
STATEMENT OF LYDIA B. PARNES, DIRECTOR, BUREAU OF CONSUMER
PROTECTION, FEDERAL TRADE COMMISSION
Ms. Parnes. Thank you, Chairman Dorgan. I appreciate the
opportunity to appear before you today to discuss the
Commission's work on issues related to online behavioral
advertising.
Balancing consumers' online privacy interests against the
development of successful online business models has been a top
priority for the Commission over the past decade. Behavioral
advertising, the use of tracking data to target advertisements
to online consumers, is a challenging issue. It may provide
benefits to consumers in the form of advertising that is more
relevant to their interests, as well as a reduction in unwanted
ads. It may also help support a diverse range of free online
content that consumers would otherwise have to pay for, for
example, blogging, search engines, social networking, and
instant access to newspapers from around the world.
At the same time, many consumers express discomfort about
the privacy and data security implications of being tracked.
Without adequate safeguards in place, consumer tracking data
could fall into the wrong hands or be used for unanticipated
purposes. These concerns are exacerbated when tracking
involves, for example, sensitive information about children's
health or a consumer's finances. Further, particular concerns
have been raised about tracking done by network advertisers
across many sites, and most recently we saw significant
consumer concern when one ISP announced and then abandoned
plans to track every move of its customers as they navigate
online.
The FTC has examined behavioral advertising for more than a
decade, almost since the Internet transformed into a commercial
medium. Our most recent efforts began in November of 2006 when
we held 3 days of public hearings on technology issues likely
to affect consumers in the next decade. Following these
hearings, Commission staff held a series of meetings with
stakeholders to learn more about behavioral advertising, and in
November 2007, the Commission hosted a town hall devoted
exclusively to behavioral advertising.
Several key points emerged at the town hall.
First, participants confirmed that online behavioral
advertising may provide valuable benefits to consumers.
Second, the invisibility of the practice to consumers
raises privacy concerns, as does the risk that data collected
for behavioral advertising could be misused.
And third, business and consumer groups alike expressed
support for transparency and consumer control in the online
marketplace.
In December 2007, following the town hall, Commission staff
issued and requested comments on proposed principles for online
behavioral advertising to spur continuing public dialogue and
encourage meaningful self-regulation. In brief, the proposed
principles identify four issues to consider in developing a
self-regulatory scheme.
First, companies that collect information for behavioral
advertising should provide meaningful disclosures to consumers
about the practice, as well as choice about whether their
information is collected for this purpose.
Second, companies should provide reasonable security for
behavioral data so that it does not fall into the wrong hands
and should retain data only as long as necessary to fulfill a
legitimate business or law enforcement need.
Third, before a company uses behavioral data in a manner
that is materially different from promises made when the data
was initially collected, it should obtain affirmative express
consent from the consumer.
Fourth, companies should obtain affirmative express consent
before they use sensitive data for behavioral advertising.
Commission staff received over 60 thoughtful, constructive,
and diverse comments on the principles. The comment period has
closed, and we are carefully evaluating the comments that we
have received. Included in the comments were a number of
specific proposals for how self-regulation could be
implemented, as well as reports about steps taken to address
privacy concerns since the town hall. Although there clearly is
more work to be done, the Commission is cautiously optimistic
that the privacy issues raised by online behavioral advertising
can be effectively addressed through self-regulation. In such a
dynamic and diverse environment, self-regulation may, indeed,
be the best means to develop workable approaches to privacy.
The Commission, of course, will continue to monitor the
marketplace to keep pace with developments, gain a better
understanding of the issues, and take appropriate action to
protect consumers as circumstances warrant.
Thank you for your attention, and I would, of course, be
happy to answer any questions.
[The prepared statement of Ms. Parnes follows:]
Prepared Statement of Lydia B. Parnes, Director,
Bureau of Consumer Protection, Federal Trade Commission
I. Introduction
Chairman Inouye, Vice Chairman Stevens, and Members of Committee, I
am Lydia Parnes,\1\ Director of the Bureau of Consumer Protection at
the Federal Trade Commission (the ``FTC'' or ``Commission''). I
appreciate the opportunity to appear before you today to discuss the
Commission's activities regarding online behavioral advertising, the
practice of collecting information about an individual's online
activities in order to serve advertisements that are tailored to that
individual's interests. Over the past year or so, the Commission has
undertaken a comprehensive effort to educate itself and the public
about this practice and its implications for consumer privacy. This
testimony will describe the Commission's efforts, which have included
hosting a ``Town Hall'' meeting and issuing for public comment FTC
staff's proposed online behavioral advertising principles.\2\
---------------------------------------------------------------------------
\1\ The views expressed in this statement represent the views of
the Commission. My oral presentation and responses to any questions are
my own, however, and do not necessarily reflect the views of the
Commission or any individual Commissioner.
\2\ See Federal Trade Commission, ``Ehavioral Advertising:
Tracking, Targeting, and Technology,'' available at http://www.ftc.gov/
bcp/workshops/ehavioral/index.shtml.
---------------------------------------------------------------------------
The Commission's examination of behavioral advertising has shown
that the issues surrounding this practice are complex, that the
business models are diverse and constantly evolving, and that
behavioral advertising may provide benefits to consumers even as it
raises concerns about consumer privacy. At this time, the Commission is
cautiously optimistic that the privacy concerns raised by behavioral
advertising can be addressed effectively by industry self-
regulation.\3\
---------------------------------------------------------------------------
\3\ Although FTC staff has proposed self-regulation to address the
general privacy concerns raised by behavioral advertising, the
Commission will of course continue to bring enforcement actions to
challenge law violations in appropriate cases.
---------------------------------------------------------------------------
II. Behavioral Advertising
Many businesses use online behavioral advertising in an attempt to
increase the effectiveness of their advertising by targeting
advertisements more closely to the interests of their audience. The
practice generally involves the use of ``cookies'' to track consumers'
activities online and associate those activities with a particular
computer or device. In many cases, the information collected is not
personally identifiable in the traditional sense--that is, the
information does not include the consumer's name, physical address, or
similar identifier that could be used to identify the consumer in the
offline world. Many of the companies engaged in behavioral advertising
are so-called ``network advertisers,'' companies that serve
advertisements across the Internet at websites that participate in
their networks.\4\
---------------------------------------------------------------------------
\4\ The advertisements are typically based upon data collected
about a given consumer as he or she travels across the different
websites in the advertising network. A website may belong to multiple
networks.
---------------------------------------------------------------------------
An example of how behavioral advertising might work is as follows:
a consumer visits a travel website and searches for airline flights to
New York City. The consumer does not purchase any tickets, but later
visits the website of a local newspaper to read about the Washington
Nationals baseball team. While on the newspaper's website, the consumer
receives an advertisement from an airline featuring flights to New York
City.
In this simple example, the travel website where the consumer
conducted his research might have an arrangement with a network
advertiser to provide advertising to its visitors. The network
advertiser places on the consumer's computer a cookie, which stores
non-personally identifiable information such as the web pages the
consumer has visited, the advertisements that the consumer has been
shown, and how frequently each advertisement has been shown. Because
the newspaper's website is also part of the advertising network, when
the consumer visits the newspaper website, the network advertiser
recognizes the cookie from the travel website as its own and identifies
the consumer as likely having an interest in traveling to New York. It
then serves the corresponding advertisement for airline flights to New
York.
In a slightly more sophisticated example, the information about the
content that the consumer had selected from the travel website could be
combined with information about the consumer's activities on the
newspaper's website. The advertisement served could then be tailored to
the consumer's interest in, not just New York City, but also baseball
(e.g., an advertisement referring to the New York Yankees).
As these examples illustrate, behavioral advertising may provide
benefits to consumers in the form of advertising that is more relevant
to their interests. Consumer research has shown that many online
consumers value more personalized ads, which may facilitate shopping
for the specific products that consumers want.\5\ Further, by providing
advertisements that are likely to be of interest to the consumer,
behavioral advertising also may reduce the number of unwanted, and
potentially unwelcome, advertisements consumers receive online.
---------------------------------------------------------------------------
\5\ See Larry Ponemon, ``FTC Presentation on Cookies and Consumer
Permissions,'' presented at the FTC's Town Hall ``Ehavioral
Advertising: Tracking, Targeting, and Technology'' (Nov. 1, 2007), at
7, available at http://www.ftc.gov/bcp/workshops/ehavioral/
presentations/31ponemon.pdf (survey found that 55 percent of
respondents believed that an online ad that targeted their individual
preferences or interests improved, to some degree, their online
experience). See also TRUSTe/TNS Presentation, TRUSTe and TNS Global,
``Consumer Attitudes about Behavioral Advertising'' at 10 (March 28,
2008) (72 percent of respondents found online advertising annoying when
it was not relevant to their interests or needs). But see infra note 13
and accompanying text.
---------------------------------------------------------------------------
More broadly, the revenue model for the Internet is, to a large
extent, advertising-based, and using behavioral techniques can increase
the cost-effectiveness of online advertising. Thus, behavioral
advertising may help subsidize and support a diverse range of free
online content and services that otherwise might not be available or
that consumers would otherwise have to pay for--content and services
such as blogging, search engines, social networking, and instant access
to newspapers and information from around the world.
At the same time, however, behavioral advertising raises consumer
privacy concerns. As described below, many consumers express discomfort
about the privacy implications of being tracked, as well as the
specific harms that could result. In particular, without adequate
safeguards in place, consumer tracking data may fall into the wrong
hands or be used for unanticipated purposes.\6\ These concerns are
exacerbated when the tracking involves sensitive information about, for
example, children, health, or a consumer's finances.
---------------------------------------------------------------------------
\6\ As a result of these concerns, a number of consumer groups and
others have asked the Commission to take action in this area. See,
e.g., Center for Digital Democracy and U.S. Public Interest Research
Group Complaint and Request for Inquiry and Injunctive Relief
Concerning Unfair and Deceptive Online Marketing Practices (Nov. 1,
2006), available at http://www.democraticmedia.org/files/pdf/
FTCadprivacy.pdf; Ari Schwartz and Alissa Cooper, Center for Democracy
and Technology, ``CDT Letter to Commissioner Rosch,'' (Jan. 19, 2007),
available at http://www.cdt.org/privacy/20070119rosch-behavioral-
letter.pdf; Mindy Bockstein, ``Letter to Chairman Majoras Re:
DoubleClick, Inc. and Google, Inc. Merger,'' New York State Consumer
Protection Board (May 1, 2007), available at http://epic.org/privacy/
ftc/google/cpb.pdf.
---------------------------------------------------------------------------
Recent high-profile incidents where tracking data has been released
have magnified consumers' concerns. In August 2006, for example, an
employee of Internet service provider and web services company AOL made
public the search records of approximately 658,000 customers.\7\ The
search records were not identified by name, and, in fact, the company
had taken steps to anonymize the data. By combining the highly
particularized and often personal searches, however, several
newspapers, including the New York Times,\8\ and consumer groups were
able to identify some individual AOL users and their queries,
challenging traditional notions about what data is or is not personally
identifiable.
---------------------------------------------------------------------------
\7\ See, e.g., Jeremy Kirk, ``AOL Search Data Reportedly
Released,'' Computerworld (Aug. 6, 2007), available at http://
computerworld.com/action/article.do?command=viewArticleBasic&
taxonomyName=privacy&articleId=9002234&taxonomyId=84.
\8\ See Michael Barbaro and Tom Zeller, ``A Face Is Exposed for AOL
Searcher No. 4417749,'' www.nytimes.com, Aug. 9, 2006, available at
http://www.nytimes.com/2006/08/09/technology/09aol.html.
---------------------------------------------------------------------------
Another incident involved the social networking site Facebook. In
November 2007, Facebook released a program called Beacon, which allowed
users to share information about their online activities, such as the
purchases they had made or the videos they had viewed. The Beacon
service tracked the activities of logged-in users on websites that had
partnered with Facebook. If a user did not opt out of this tracking,
Facebook's partner sites would send to Facebook information about the
user's purchases at the partner sites. Facebook then published this
information on the user's profile page and sent it to the user's
Facebook ``friends.''
The Beacon program raised significant concerns among Facebook
users.\9\ Approximately 30 groups formed on Facebook to protest Beacon,
with one of the groups representing over 4,700 members,\10\ and over
50,000 Facebook users signed a petition objecting to the new
program.\11\ Within a few weeks, Facebook changed its program by adding
more user controls over what information is shared with ``friends'' and
by improving notifications to users before sharing their information
with others on Facebook.\12\
---------------------------------------------------------------------------
\9\ In one now-famous example, a man had bought a ring for his wife
as a surprise; the surprise was ruined when his wife read about his
purchase on the man's user profile page. See, e.g.,
Ellen Nakashima, ``Feeling Betrayed, Facebook Users Force Site to Honor
Privacy,'' Washingtonpost.com, (Nov. 30, 2007), available at http://
www.washingtonpost.com/wp-dyn/content/article/2007/11/29/
AR2007112902503_pf.html.
\10\ See Facebook home page, http://www.facebook.com, viewed on
March 21, 2008.
\11\ MoveOn.org Civic ActionTM created an online
petition for consumers to express their objection to Facebook's Beacon
program. The petition stated, ``Sites like Facebook must respect my
privacy. They should not tell my friends what I buy on other sites--or
let companies use my name to endorse their products--without my
explicit permission.'' MoveOn.org Civic Action Petition, available at
http://www.civic.moveon.org/facebookprivacy/, viewed June 9, 2008.
\12\ See Reuters News, ``Facebook Makes Tweak After Privacy
Protest,'' RedHerring.com, Nov. 30, 2007, available at http://
www.redherring.com/Home/23224.
---------------------------------------------------------------------------
Surveys confirm that consumers are concerned about the privacy of
their activities as they navigate online. For example, in two recent
surveys, a majority of consumers expressed some degree of discomfort
with having information about their online activities collected and
used to serve advertising.\13\ Similarly, only 20 percent of consumers
in a third survey stated that they would allow a marketer to share
information about them in order to track their purchasing behaviors and
to help predict future purchasing decisions.\14\ Another survey found
that 45 percent of consumers believe that online tracking should be
banned, and another 47 percent would allow such tracking, but only with
some form of consumer control.\15\ These surveys underscore the
importance of online privacy to consumers and highlight the fundamental
importance of maintaining trust in the online marketplace.
---------------------------------------------------------------------------
\13\ See Alan Westin, ``Online Users, Behavioral Marketing and
Privacy: Results of a National Harris/Westin Survey'' (March 2008)
(almost 60 percent of respondents were ``not comfortable'' to some
degree with online behavioral marketing); TRUSTe/TNS Presentation,
``Behavioral Advertising: Privacy, Consumer Attitudes and Best
Practices,'' at 10 (April 23, 2008) (57 percent of respondents were not
comfortable with advertisers using browsing history to serve ads, even
if the information is not connected to personally identifiable
information).
\14\ See Ponemon Presentation, supra note 5, at 11.
\15\ See George R. Milne, ``Information Exchange Expectations of
Consumers, Marketing Managers and Direct Marketers,'' University of
Massachusetts Amherst (presented on Nov. 1, 2007), available at http://
www.ftc.gov/bcp/workshops/ehavioral/presentations/3gmilne.pdf.
---------------------------------------------------------------------------
III. FTC Initiatives Concerning Consumer Privacy and Behavioral
Advertising
Since privacy first emerged as a significant consumer protection
issue in the mid-1990s, it has been one of the Commission's highest
priorities. The Commission has worked to address privacy issues through
consumer and business education, law enforcement, and policy
initiatives. For example, the FTC has promulgated and enforced the Do
Not Call Rule to respond to consumer complaints about unsolicited and
unwanted telemarketing; \16\ has waged a multi-faceted war on identity
theft; \17\ has encouraged better data security practices by businesses
through educational initiatives \18\ and a robust enforcement program;
\19\ has brought numerous enforcement actions to reduce the incidence
of spun and spyware; \20\ and has held numerous workshops to examine
emerging technologies and business practices, and the privacy and other
issues they raise for consumers.\21\ In early 2006, recognizing the
ever-increasing importance of privacy to consumers and to a healthy
marketplace, the Commission established the Division of Privacy and
Identity Protection, a division devoted exclusively to privacy-related
issues.
---------------------------------------------------------------------------
\16\ Telemarketing Sales Rule: Final Rule, 16 C.F.R. Part 310
(2003), available at http://www.ftc.gov/os/2003/01/tsrfrn.pdf.
\17\ See, e.g., FTC ID theft website, available at www.ftc.gov/
idtheft. In one recent effort, the FTC coordinated with the U.S. Postal
Service to send a letter to every American household containing
information about how to protect against identity theft. See Press
Release, ``Postmaster General Sends Advice to Prevent ID Theft,'' U.S.
Postal Service (Feb. 19, 2008), available at http://www.usps.com/
connnunications/newsroom/2008/pr08_014.htm.
\18\ See, e.g., Federal Trade Commission, ``Protecting Personal
Information: A Guide for Business,'' available at http://www.ftc.gov/
infosecurity/; see also http://onguardonline.gov/index.html.
\19\ Since 2001, the Commission has obtained twenty consent orders
against companies that allegedly failed to provide reasonable
protections for sensitive consumer information. See In the Matter of
The TJX Companies, FTC File No. 072-3055 (Mar. 27, 2008, settlement
accepted for public comment); In the Matter of Reed Elsevier Inc. and
Seisint Inc., FTC File No. 052-3094 (Mar. 27, 2008, settlement accepted
for public comment); United States v. ValueClick, Inc., No. CV08-01711
(C.D. Cal. Mar. 13, 2008); In the Matter of Goal Financial, LLC, FTC
Docket No. C-4216 (April 15, 2008); In the Matter of Life is Good,
Inc., FTC Docket No. C-4218 (Apr. 18, 2008); United States v. American
United Mortgage, No. CV07C 7064, (N.D. Ill. Dec. 18, 2007); In the
Matter of Guidance Software, Inc., FTC Docket No. C-4187 (Apr. 3,
2007); In the Matter of CardSystems Solutions, Inc., FTC Docket No. C-
4168 (Sept. 5, 2006); In the Matter of Nations Title Agency, Inc., FTC
Docket No. C-4161 (June 19, 2006); In the Matter of DSW, Inc., FTC
Docket No. C-4157 (Mar. 7, 2006); United States v. ChoicePoint, Inc.,
No. 106-CV-0198 (N.D. Ga. Feb. 15, 2006); In the Matter of Superior
Mortgage Corp., FTC Docket No. C-4153 (Dec. 14, 2005); In the Matter of
BJ's Wholesale Club, Inc., FTC Docket No. C-4148 (Sept. 20, 2005); In
the Matter of Nationwide Mortgage Group, Inc., FTC Docket No. 9319
(Apr. 12, 2005); In the Matter of Petco Animal Supplies, Inc., FTC
Docket No. C-4133 (Mar. 4, 2005); In the Matter of Sunbelt Lending
Services, FTC Docket No. C-4129 (Jan. 3, 2005); In the Matter of MTS
Inc., d/b/a Tower Records/Books/Video, FTC Docket No. C-4110 (May 28,
2004); In the Matter of Guess?, Inc., FTC Docket No. C-4091 (July 30,
2003); In the Matter of Microsoft Corp., FTC Docket No. C-4069 (Dec.
20, 2002); In the Matter of Eli Lilly & Co., FTC Docket No. C-4047 (May
8, 2002).
\20\ Since 2004, the Commission has initiated eleven spyware-
related law enforcement actions. Detailed information regarding each of
these law enforcement actions is available at http://www.ftc.gov/bcp/
edu/microsites/spyware/law_enfor.htm. Since 1997, when the FTC brought
its first enforcement action targeting unsolicited commercial e-mail,
or ``spam,'' the FTC has brought 94 law enforcement actions. See
generally Report on ``Spam Summit: The Next Generation of Threats and
Solutions'' (Nov. 2007), available at http://www.ftc.gov/os/2007/12/
071220spamsummitreport.pdf.
\21\ See discussion infra pp. 9-12.
---------------------------------------------------------------------------
In developing and implementing its privacy program, the FTC has
been mindful of the need for flexibility and balance--that is, the need
to address consumer concerns and harms without stifling innovation or
imposing needless costs on consumers and businesses.
A. 1999 Workshop on Online Profiling
The Commission first examined the issue of behavioral advertising
in 1999, when it held a joint public workshop with the Department of
Commerce on the practice--then called ``online profiling.'' The
workshop examined the practice of tracking consumers' activities
online, as well as the role of self-regulation in this area.
In response to the concerns highlighted at the workshop, industry
members formed the Network Advertising Initiative (``NAI''), a self-
regulatory organization addressing behavioral advertising by network
advertisers. Shortly thereafter, the NAI issued the NAI Self-Regulatory
Principles (``NAI Principles'') governing collection of information for
online advertising by network advertisers.\22\ In the early 2000s,
however, with the ``burst'' of the dot com bubble, many network
advertisers--including most of the NAI membership--went out of
business.
---------------------------------------------------------------------------
\22\ Briefly, the NAI Principles set forth guidelines for online
network advertisers and provide a means by which consumers can opt out
of behavioral advertising at a centralized website. For more
information on the FTC workshop and NAI, see Online Profiling: A Report
to Congress (June 2000) at 22 and Online Profiling: A Report Congress
Part 2 Recommendations (July 2000), available at http://www.ftc.gov/os/
2000/06/onlineprofilingreportjune2000.pdf and http://
www.networkadvertising.org. As discussed further below, NAI recently
proposed for public comment revised NAI Principles.
---------------------------------------------------------------------------
Emblematic of the highly dynamic nature of the online environment,
by the time the FTC held its public hearings on Protecting Consumers in
the Next Tech-ade (``Tech-ade'') only a few years later,\23\ the issue
of online tracking and advertising had reemerged. In the intervening
years, behavioral advertising had become a highly successful business
practice, and a number of Tech-ade participants raised concerns about
its effects on consumer privacy.
---------------------------------------------------------------------------
\23\ The purpose of the Tech-ade hearings, held in November 2006,
was to examine the technological and consumer protection developments
anticipated over the next decade. See generally http://www.ftc.gov/bcp/
workshops/techade/index.html.
---------------------------------------------------------------------------
B. The FTC Town Hall on Online Behavioral Advertising
Beginning in Fall 2006, the Commission staff held a series of
meetings with numerous industry representatives, technology experts,
consumer and privacy advocates, and academics to learn more about the
practice of behavioral advertising. The purpose of these meetings was
to explore further the issues raised at Tech-ade, learn about
developments since the FTC's 1999 Workshop, and examine concerns about
behavioral advertising that had been raised by privacy advocates and
others.\24\ Seeking a broader forum in which to examine and discuss
these issues, and particularly the privacy issues raised by the
practice, the FTC held a two-day Town Hall meeting on behavioral
advertising in November 2007.
---------------------------------------------------------------------------
\24\ See CDD et al., Complaint and Request for Inquiry and
Injunctive Relief, supra note 6. Many of these concerns were amplified
by the announcement of the proposed merger between Google and
DoubleClick in April 2007. The Commission approved the merger on
December 20, 2007, at the same time that it issued FTC staff's proposed
self-regulatory guidelines. See ``Staff Proposes Online Behavioral
Advertising Policy Principles,'' Federal Trade Commission (Dec. 20,
2008), available at http://www.ftc.gov/opa/2007/12/principles.shtm. The
Principles are discussed infra at 13.
---------------------------------------------------------------------------
From the Town Hall, as well as the meetings preceding it, several
key points emerged. First, as discussed above, online behavioral
advertising may provide many valuable benefits to consumers in the form
of free content, personalization that many consumers value, and a
potential reduction in unwanted advertising. Second, the invisibility
of the practice to consumers raises privacy concerns, as does the risk
that data collected for behavioral advertising--including sensitive
data about children, health, or finances--could be misused. Third,
business and consumer groups alike expressed support for transparency
and consumer control in the online marketplace.
Many participants at the Town Hall also criticized the self-
regulatory efforts that had been implemented to date. In particular,
these participants stated that the NAI Principles had not been
sufficiently effective in addressing the privacy concerns raised by
behavioral advertising because of the NAP s limited membership, the
limited scope of the NAI Principles (which apply to network advertisers
but not to other companies engaged in behavioral advertising), and the
NAI Principles' lack of enforcement and cumbersome opt-out system.\25\
Further, while other industry associations had promulgated online self-
regulatory schemes to address privacy issues, these schemes had not
generally focused on behavioral advertising.\26\
---------------------------------------------------------------------------
\25\ According to critics, the NAI Principles' opt-out mechanism is
difficult to locate and use because it is located on the NAI website,
where consumers would be unlikely to find it. As noted above, in April
of this year, the NAI issued a proposed revised set of self-regulatory
principles designed to address criticisms of the original NAI
Principles and to respond to the FTC staff's call for stronger self-
regulation. The NAI has sought comment on its proposed revised
principles, and comments were due June 12, 2008. See ``Self-Regulatory
Principles for Online Preference Marketing By Network Advertisers,''
Network Advertising Initiative (issued April 10, 2008), available at
http://www.networkadvertising.org/pdfs/NAI_principles.pdf.
\26\ Since the Town Hall, some of these industry groups, as well as
several online companies and privacy groups, have sought to address the
concerns raised about behavioral advertising. See, e.g., Interactive
Advertising Bureau, ``Privacy Principles,'' (adopted Feb. 24, 2008),
available at http://www.iab net/iab_products_and_industry_services/
1421/1443/1464; Comment ``Online Behavioral Advertising: Moving the
Discussion Forward to Possible Self-Regulatory Principles,'' Microsoft
Corp. (April 11, 2008), available at http://www.ftc.gov/os/comments/
behavioraladprinciples/080411microsoft.pdf; Comment ``FTC Staff
Proposed Online Behavioral Advertising Principles: Comments of AOL,
LLC,'' AOL, LLC (April 11, 2008), available at http://www.ftc.gov/os/
comments/behavioraladprinciples/080411aol.pdf; Ari Schwartz, Center for
Democracy and Technology, et al., ``Consumer Rights and Protections in
the Behavioral Advertising Sector,'' (Oct. 31, 2007) (proposing a ``Do
Not Track List'' designed to increase consumers' control over tracking
of their activities online), available at http://www.cdt.org/privacy/
20071031consumerprotectionsbehavioral.pdf.
---------------------------------------------------------------------------
C. The FTC's Proposed Self-Regulatory Principles
In December 2007, in response to the issues discussed at the Town
Hall and in public comments received in connection with that event,
Commission staff issued and requested comment on a set of proposed
principles titled, ``Behavioral Advertising: Moving the Discussion
Forward to Possible Self-Regulatory Principles'' (the ``Principles'').
The proposed Principles address the central concerns about online
behavioral advertising expressed by interested parties; they also build
upon existing ``best practices'' in the area of privacy, as well as (in
some cases) previous FTC guidance and/or law enforcement actions. At
the same time, the Principles reflect FTC staff's recognition of the
potential benefits provided by online behavioral advertising and the
need to maintain vigorous competition in this area.
The purpose of the proposed Principles is to encourage more
meaningful and enforceable self-regulation. At this time, the
Commission believes that self-regulation may be the preferable approach
for this dynamic marketplace because it affords the flexibility that is
needed as business models continue to evolve.
In brief, the staff proposal identifies four governing principles
for behavioral advertising.\27\ The first is transparency and consumer
control: companies that collect information for behavioral advertising
should provide meaningful disclosures to consumers about the practices,
as well as choice about whether their information is collected for this
purpose.\28\ The second principle is reasonable security: companies
should provide reasonable security for behavioral data so that it does
not fall into the wrong hands, and should retain data only as long as
necessary to fulfill a legitimate business or law enforcement need.\29\
The third principle governs material changes to privacy policies:
before a company uses behavioral data in a manner that is materially
different from promises made when the data was collected, it should
obtain affirmative express consent from the consumer.\30\ This
principle ensures that consumers can rely on promises made about how
their information will be used, and can prevent contrary uses if they
so choose. The fourth principle states that companies should obtain
affirmative express consent before they use sensitive data--for
example, data about children, health, or finances--for behavioral
advertising.\31\
---------------------------------------------------------------------------
\27\ Recent news reports have highlighted concerns about behavioral
advertising involving Internet Service Providers (``ISPs''). The ISP-
based model for delivering behaviorally-targeted advertising may raise
heightened privacy concerns because it could involve the tracking of
subscribers wherever they go online and the accumulation of vast stores
of data about their online activities. Further, information about the
subscriber's activities potentially could be combined with the
personally identifiable information that ISPs possess about their
subscribers. In issuing the proposed Principles for public comment, FTC
staff intended the Principles to apply to ISPs.
\28\ For more information and guidance on the use of disclosures in
online advertising, see Dot Com Disclosures, Information About Online
Advertising, http://www.ftc.gov/bcp/conline/pubs/buspubs/dotcom/
index.shtml (May 2000).
\29\ The FTC has highlighted the need for reasonable security in
numerous educational materials and enforcement actions to date. See
supra notes 18-19.
\30\ See, e.g., Gateway Learning Corp., Docket No. C-4120 (Sept.
10, 2004), http://www.ftc.gov/opa/2004/07/gateway.shtm (company made
material changes to its privacy policy and allegedly applied such
changes to data collected under the old policy; opt-in consent required
for future such changes).
\31\ Commission staff also sought comment on the potential uses of
tracking data beyond behavioral advertising.
---------------------------------------------------------------------------
IV. Next Steps
In response to the request for public comment, Commission staff
received over 60 comments on the Principles, representing many
thoughtful and constructive views from diverse business sectors,
industry self-regulatory bodies, privacy advocates, technologists,
academics, and consumers. The comment period for the Principles has
closed, and Commission staff is carefully evaluating the comments
received.
Included in the comments were a number of specific proposals for
how self-regulation could be implemented, as well as reports regarding
steps taken to address privacy concerns since the Town Hall. The FTC is
encouraged by the efforts that have already been made by the NAI \32\
and some other organizations and companies \33\ and believes that the
self-regulatory process that has been initiated is a promising one.
Although there is more work to be done in this area, the Commission is
cautiously optimistic that the privacy issues raised by online
behavioral advertising can be effectively addressed through meaningful,
enforceable self-regulation. The dynamic and diverse online environment
demands workable and adaptable approaches to privacy that will be
responsive to the evolving marketplace. Nevertheless, the Commission
will continue to closely monitor the marketplace so that it can take
appropriate action to protect consumers as the circumstances warrant.
---------------------------------------------------------------------------
\32\ Current NAI members include DoubleClick, Yahoo! Inc., TACODA,
Inc., Acerno, AlmondNet, BlueLithium, Mindset Media, Revenue Science,
Inc., 24/7 Real Media Inc., and Undertone Networks.
\33\ See supra note 26. Although many organizations and consumer
groups have undertaken efforts to address FTC staff's proposed
Principles, a few organizations have expressed concern that
implementing the Principles would be too costly and would undermine
continued development of the online marketplace. FTC staff is
evaluating all of these comments as it considers next steps in this
area.
---------------------------------------------------------------------------
V. Conclusion
The Commission appreciates this opportunity to discuss its work on
behavioral advertising. The Commission is committed to addressing new
and emerging privacy issues such as online behavioral advertising and
looks forward to working further with the Committee on this important
consumer issue.
Senator Dorgan. Ms. Parnes, thank you very much.
Next we will hear from Ms. Jane Horvath, the Senior Privacy
Counsel at Google, Incorporated. Ms. Horvath?
STATEMENT OF JANE HORVATH, SENIOR PRIVACY COUNSEL, GOOGLE, INC.
Ms. Horvath. Senator Dorgan and Senator Vitter, the most
important point I would like to make this morning is simple.
Google makes privacy a priority because our business depends on
it. If our users are uncomfortable with how we manage their
personal information, they are only one click away from
switching to a competitor's services.
Putting our users first means that we are deeply committed
to their privacy, and succeeding in online advertising and
protecting our users' privacy are not mutually exclusive goals.
This morning I will first discuss how online advertising
benefits advertisers, website publishers, and Internet users.
Second, I will discuss Google's approach to privacy. And
finally, I will make recommendations for how government and
industry can better protect Internet users' privacy.
So let me first touch on the benefits of online
advertising. Google's two primary advertising programs, AdWords
and AdSense, provide users with highly relevant ads, match
advertisers with users who are interested in their products,
and provide revenue for website publishers who place our ads on
their sites. For example, in Minneapolis, taxi driver Kenny
Kormendy, built a website for out-of-state travelers called
Gopher State Taxi and used Google's AdWords program to compete
online with bigger taxi companies. Today Gopher State Taxi has
grown to a network of over 36 cabs, and Mr. Kormendy credits
Google with connecting 9 out of 10 of its customers.
When someone clicks on one of our ads on a website, Google
also shares revenue from that ad with the website owner. Last
year we paid a total of $4.5 billion in ad revenue to website
publishers across the United States.
Next, let me talk about Google's approach to privacy. As I
said earlier, Google makes privacy a priority because our
business depends on it. We make sure that three design
fundamentals are the bedrock of our privacy practices.
First, transparency. We have been an industry leader in
finding new ways to educate users about privacy such as through
our Google Privacy Channel on YouTube where we feature videos
that explain our privacy policies in plain language.
Second, choice. We strive to design our products in a way
that gives users meaningful choices about what information they
provide to us. For example, our Google Talk instant messaging
service includes an ``off the record'' feature that prevents
either party from storing the chat.
And third, security. We take seriously the protection of
data that our users entrust with us. Google employs some of the
world's best engineers in software and network security and has
teams dedicated to developing information safeguards. Google's
advertising products are primarily driven by context rather
than behavior. Unlike other companies we have built our
business on showing ads that are relevant to what a user is
looking for, not by building detailed profiles based on a
user's online behavior.
As we continue to incorporate DoubleClick's display ad
serving capabilities into our business, Google will continue to
be a leader in offering products that respect privacy.
Finally, let me turn to our efforts to innovate in the area
of privacy protection. Feedback from our users and outside
parties, as well as our own internal discussions, has led us to
several privacy innovations, including our decision last year
to anonymize our server logs after 18 months. In that spirit of
innovation today, we offer the following recommendations for
both government and the private sector.
First, Google supports the passage of a comprehensive
Federal privacy law that will establish a uniform framework for
privacy and put penalties in place to punish and dissuade bad
actors.
Second, we support the Federal Trade Commission's efforts
to develop principles relating to online privacy and behavioral
advertising, and we hope that revised principles will be
adopted widely by the online ad industry.
And third, we believe that greater labeling of online
display ads should be adopted as an industry standard.
As I conclude my testimony this morning and welcome the
Committee's questions, I would like to show a brief excerpt
from one of the videos on our Google Privacy YouTube Channel.
This video shows a user how to easily remove cookies from their
web browsers. Thank you.
[Video shown.]
[The prepared statement of Ms. Horvath follows:]
Prepared Statement of Jane Horvath, Senior Privacy Counsel, Google,
Inc. Chairman Inouye, Vice Chairman Stevens, Members of the Committee.
I'm pleased to appear before you this morning to discuss online
advertising and the ways that Google protects our users' privacy. My
name is Jane Horvath, and I am Google's Senior Privacy Counsel. In that
role I am responsible for working with our product teams and other
privacy professionals at Google to ensure compliance with privacy laws
and develop best practices for protecting our users' privacy.
Google's mission is to organize the world's information and make it
universally accessible and useful. The best known way that we do this
today is through our search engine, which is available for free to
Internet users throughout the world. The availability of Google search
and our other products--and the improvements that we make to our
products on a daily basis--is funded by online advertising, by far our
primary source of revenue.
Online advertising is relatively young and a very small piece of
the advertising market as a whole. It is a dynamic business
characterized by strong competition, significant innovation, and
continuing growth. Online advertising has succeeded because it helps
businesses find customers more efficiently and effectively than through
other media. It has also helped to create entirely new and innovative
small businesses that generate revenue through advertising, often in
partnership with Google.
At Google we believe that our online advertising business has
succeeded because our most important advertising goal is to deliver ads
that benefit our users. From its inception, Google has focused on
providing the best user experience possible. We do this, for example,
by ensuring that advertising on our site delivers relevant content that
is not a distraction. In fact, our goal is to make our ads just as
useful to Google's users as search results themselves.
We've also made a commitment to never compromise the integrity of
our search results, for example by manipulating rankings to place our
partners higher in our search results. And advertising on Google is
always clearly identified as a ``Sponsored Link'' to ensure that our
users know the difference between our search results and any
advertising that we provide.
Putting our users first also means that we are deeply committed to
their privacy, and our products and policies demonstrate that
commitment. We believe that success in online advertising and
protecting our users' privacy are not mutually exclusive goals. We work
hard to provide advertising in a way that is transparent to users,
provides them with appropriate choices, and protects any personal
information that we collect from inappropriate access by third parties.
In my testimony this morning, I would like to cover three key
points:
First, I'll explain Google's main advertising products and the
significant benefits that we at Google believe online
advertising brings to advertisers, online publishers, and
individual Internet users.
Second, I'll discuss Google's approach to privacy, specific
steps that we take to protect our users' privacy, and privacy
issues involving our advertising business.
And finally, I'll explore ideas and make recommendations for
how to better protect Internet users' privacy both with respect
to advertising as well as more generally as more and more
information moves to the Internet cloud.
The Benefits of Online Advertising
Google offers three main advertising products: AdWords, AdSense for
Search, and AdSense for Content. Our AdWords product allows us to
provide ads on Google.com in response to search queries entered by our
users, as well as to provide ads on our AdSense for Content and AdSense
for Search services. AdSense for Search allows us to provide ads in
response to search queries entered by users of our partners' search
engines, including AOL and Ask.com. AdSense for Content allows us to
provide ads to visitors of our third-party publisher partners'
websites. AdSense for Content ads are provided based on the content of
the page that is being viewed by a user. The vast majority of the
revenue that Google generates comes from these three products.
All three advertising products are primarily easy-to-create text
ads, which is one of the many reasons that hundreds of thousands of
small businesses advertise with us. We also provide the capability to
show display ads--ads that incorporate graphics in addition to text--
through AdSense for Content, and we plan to enhance our display ad
serving capabilities with our recent acquisition of DoubleClick, a
display ad serving technology company.
Advertisers, online publishers, and consumers all benefit from our
advertising network. I'll start with consumers--our users--on whom our
business depends.
In our experience, users value the advertisements that we deliver
along with search results and other web content because the ads help
connect them to the information, products, and services they seek. The
ads we deliver to our users complement the natural search results that
we provide because our users are often searching for products and
services that our advertisers offer. Making this connection is
critical, and we strive to deliver the ads that are the most relevant
to our users, not just the ones that generate the most revenue for us.
We do this through our innovative ad auction system, which gives weight
to the relevancy--the usefulness--of the ad to our users based on their
search queries or the content that they are viewing. And in our pay-
per-click pricing model we only generate revenue when a user is
interested enough to click on an ad.
The revenue that we generate from online advertising makes it
possible for Google to offer dozens of free products to our users--
everything from search and e-mail to our word processing application,
Google Docs. Each of these products underscores our commitment to
improving our users' online experience. For example, Google Docs allows
multiple users to edit a single document, presentation, or spreadsheet
at the same time. And, despite the popularity of tools like Google
Earth and YouTube, each of our products is free to individuals for
personal use. Our online advertising business model subsidizes the
creation, development, and ongoing improvements to and support for
these and future products.
And our ads aren't always commercial. We run a program called
Google Grants that provides free advertising to not-for-profit
organizations engaged in areas such as science and technology,
education, global public health, the environment, youth advocacy, and
the arts. For example, we have provided Google Grants to non-profits
such as Room to Read (www.roomtoread.org), which educates children in
Vietnam, Nepal, India, and Cambodia, and CoachArt (www.coachart.org),
which provides therapeutic art and athletic lessons to underprivileged
children with life-threatening illnesses. Since April 2003, our
grantees have collectively received almost $300 million in free
advertising.
Our advertising network also enables small businesses to connect
with consumers that they otherwise would not reach, and to do so
affordably, efficiently, and effectively. The advertiser decides the
maximum amount of money it wishes to spend on advertising and, as noted
above, in the cost-per-click payment model the advertiser only pays
Google when a user actually clicks on an ad.
Here are just two of many stories of small businesses succeeding
thanks to Google advertising. Suzanne Golter owns the Happy Hound dog
daycare (www.happyhound.com) in Oakland, California. She estimates that
90 percent of her business is generated through Google AdWords, which
helps her bring in approximately 40 new clients per month. In
Minneapolis, Minnesota, Kenny Kormendy, a then-struggling taxi driver
built a site for out-of-state travelers called Gopher State Taxi
(www.gopherstatetaxi.com) and utilized AdWords to compete online with
bigger taxi companies. In under 3 years, Gopher State Taxi has grown to
a network of over 36 cabs, and Mr. Kormendy credits AdWords with
connecting nine out of ten customers that his company services.
Online advertising also promotes freer, more robust, and more
diverse speech. It's no coincidence that blogs have proliferated over
the past few years. Our AdSense product enables bloggers and other
publishers to generate revenue from ads that we place on their
websites. Without online advertising, the individuals who run these
sites would not be able to dedicate as much time and attention to their
publications as they do today. In fact, we know that many website
owners can afford to dedicate themselves to their sites full time
because of online advertising.
AdSense revenues support hundreds of thousands of diverse websites,
and a significant percentage of the revenue we earn from advertising
ends up in the hands of the bloggers and website operators who partner
with us by featuring ads provided by Google. Last year we paid $4.5
billion in advertising revenue from our AdSense program to our
publishing partners. In Nevada, Arizona, Florida, and Washington alone
over 100,000 of our publishing partners collectively generated nearly
$100 million from AdSense in 2007.
The vast majority of these AdSense partners are small businesses.
For example, in Oregon, Hope Pryor, a grandmother of four, uses AdSense
on her site--Cooksrecipes.com--to generate her primary source of
income. And in Massachusetts, honey bee aficionado and retiree Albert
Needham uses AdSense revenue generated from his Bees-online.com website
to fund personal vacations. Similar small business success stories are
found all across the United States.
It's no mistake that I've focused mainly on individual users, small
publishers, and small advertisers. Google's business model has
concentrated on what's known as the ``long tail'' of the Internet--the
millions of individuals and small businesses that cater to and need to
connect with niche interests and markets. Google's advertising programs
lower the barrier to entry for small publishers and advertisers alike,
and connect them with users who are interested in what they have to say
or sell. As our advertising business continues to grow and evolve, we
will continue working hard to encourage the development of the long
tail.
Google and Privacy
We believe user trust is essential to building the best possible
products. With every Google product, we work hard to earn and keep that
trust with a long-standing commitment to protect the privacy of our
users' personal information. We make privacy a priority because our
business depends on it. In fact, if our users are uncomfortable with
how we manage their personal information, they are only one click away
from switching to a competitor's services.
Because user trust is so critical to us, we've ensured that privacy
considerations are deeply embedded in our culture. Though I am Google's
Senior Privacy Counsel, I am just one of many individuals at Google who
work on privacy. For example, we have product counsels who work with
engineers and product managers from the beginning of product
development to ensure that our products protect our users' privacy. We
also have product managers dedicated to privacy and other trust and
safety issues. And we have a Privacy Council, which is comprised of a
cross-functional group of Google employees that convenes on a regular
basis to help Google address privacy issues.
Google's focus on user trust and privacy means that our product
teams are thinking about user privacy by building privacy protections
into our products from the ground up. For example, we have designed
most of our products to allow people to use them anonymously, and to
ensure that none of our products use any personally identifiable data
unless that use is fully disclosed in our privacy policy.
We have also made sure that three design fundamentals--all of them
rooted in fair information principles--are at the bedrock of our
privacy products and practices:
Transparency: We believe in being upfront with our users
about what information we collect and how we use it so that
they can make informed choices about their personal
information. We have been an industry leader in finding new
ways to educate users about privacy, such as through our Google
Privacy Channel on YouTube (found at www.youtube.com/
googleprivacy) where we feature privacy videos that explain our
privacy policies, practices, and product features in simple,
plain language.
Choice: We strive to design our products in a way that gives
users meaningful choices about how they use our services and
what information they provide to us. Many of our products,
including our Search service, do not require users to provide
any personally identifying information at all. When we do ask
for personal information, we also endeavor to provide features
that give users control over that information. For example, our
Google Talk instant messaging service includes an ``off the
record'' feature that prevents either party from storing the
chat.
Security: We take seriously the protection of data that our
users entrust with us. Google employs some of the world's best
engineers in software and network security and has teams
dedicated to developing and implementing policies, practices
and technologies to protect this information. More information
about our approach to security can be found in a recent post at
the Official Google Blog located at googleblog.blogspot.com/
2008/03/how-google-keeps-your-information.html.
One of our newest products is Google Health, which enables
individuals to consolidate and store their medical records and personal
health information online. Google Health demonstrates our commitment to
all three design fundamentals. For example, we have provided
significant transparency about Google Health's privacy features through
blog posts and the product's easy-to-understand privacy policy and
frequently asked questions. In addition, Google Health provides users
choice by empowering them with the decision of what information to
import, share, and delete, and easy tools for accomplishing each.
The online advertising products that we offer today are also
privacy-friendly because they are primarily contextual in nature. That
is, we generally provide ads in response to what a user is searching
for or viewing at the time, rather than based on who we believe the
user may be or an extended history of the user's activities either
online or off.
To respond to our users' desire for more relevant advertising, and
to advertisers' desire to provide more relevant advertising to Internet
users, we are experimenting with some forms of online advertising that
do involve more than the current search query to provide an ad. For
example, we are currently experimenting in Google.com search with
providing ads based on both the current query and a previous search. A
user who types ``Italy vacation'' into the Google search box, for
instance, might see ads about Tuscany or affordable flights to Rome. If
the user were to subsequently search for ``weather,'' we might assume
that there is a link between ``Italy vacation'' and ``weather'' and
deliver ads regarding local weather conditions in Italy. However,
Google does not build a profile of the user to serve these ads that is
stored and used later to serve other ads to the user.
As we continue to incorporate DoubleClick into our business, our
focus on display advertising--ads that feature images in addition to
text--will increase across our advertising product offerings, as will
our ability to provide metrics and an improved user experience to our
AdSense network. We believe that expanding into display advertising
products is one way that we can compete effectively in the highly
competitive online advertising environment. This transition will not
undermine Google's focus on privacy or our commitment to the
fundamental principles of transparency, choice, and security. As we
move to offer more display advertising and other advertising products,
Google intends to continue to be a leader in offering products that
protect and respect the privacy of our users.
Google's Efforts to Continue Innovating in Privacy
In our quickly evolving business environment, ensuring that we earn
and keep our users' trust is an essential constant for building the
best possible products. With every Google product, we work hard to earn
and keep that trust with a long-standing commitment to protect the
privacy of our users' personal information. As stated above, the
bedrock of our privacy practices are three design fundamentals:
transparency, choice, and security.
Another constant that we have found in our business is that
innovation is a critical part of our approach to privacy. To best
innovate in privacy, we welcome the feedback of privacy advocates,
government experts, our users, and other stakeholders. This feedback,
and our own internal discussions about how to protect privacy, has led
us to several privacy innovations including our decision last year to
anonymize our server logs after 18 months.
In the interest of continuing to protect individuals' privacy, we
offer the following policy and technology recommendations--some of
which can be accomplished by the private sector and some of which
involve a government role--in the spirit of continuing the effort to
innovate on consumer privacy. Our ideas and recommendations endorse a
baseline and robust level of privacy protections for all individuals.
On top of that baseline platform we believe that the private sector and
government should cooperate to educate and inform consumers about
privacy issues and to establish best practices that will help guide the
development of the quickly evolving and innovative online advertising
space. Finally, we believe that Google and others in the online
advertising industry should work to provide tools to better protect
individuals' privacy, and that government should encourage companies to
experiment with new and innovative ways of protecting consumers'
privacy.
Comprehensive Federal Privacy Law
Google supports the passage of a comprehensive Federal privacy law
that would accomplish several goals such as building consumer trust and
protections; establishing a uniform framework for privacy, which would
create consistent levels of privacy from one jurisdiction to another;
and putting penalties in place to punish and dissuade bad actors. We
believe that as information flows increase and more and more
information is processed and stored in the Internet cloud--on remote
servers rather than on users' home computers--there is a greater need
for uniform data safeguarding standards, data breach notification
procedures, and stronger procedural protections relating to government
and third party litigant access to individuals' information.
Behavioral Advertising Principles
We have participated actively in the Federal Trade Commission's
efforts to develop privacy principles relating to online privacy and
behavioral advertising. Our hope is that revised principles will be
adopted widely by the online advertising industry and serve as a model
for industry self-regulation in jurisdictions beyond the United States.
In order for the principles to achieve such broad adoption, however,
they need to be revised to ensure that they can be operationalized by
industry and that they will give consumers appropriate transparency,
choice, and security. In order for that to happen, the principles
would, among other things, need to make a distinction between
personally identifiable information (PII) and non-PII.
Consumer Education
Transparency is one of Google's bedrock design principles because
we believe that informed and knowledgeable users are best able to
protect their privacy. We believe that both the private sector and the
government, including agencies like the FTC, can and should provide
more information about what kinds of personal information are collected
by companies, how such data is used, and what steps consumers can take
to better protect their privacy.
At Google, for example, we take great pride in our effort to
provide our users with a better understanding of how we collect, use,
and protect their data through a series of short videos available at
Google.com and on YouTube, as well as through blog posts. Too often,
website operators view their online privacy policy--which is typically
impenetrable to the average user--as the beginning and end of their
privacy obligations. Web companies that interact with individuals need
to do more than simply provide and link to privacy policies; we need to
offer consumer-friendly materials in different media to better help
their users understand how their information is collected and used, and
what choices they have to protect their privacy.
Transparency and Choice in Display Advertising
Google text ads are generally labeled ``Ads by Google'' or
``Sponsored Links'' and are accompanied by an explanation of what they
are so that users understand that they are advertisements and that they
have been provided by Google. We believe that this kind of notice and
explanation should be adopted by industry and applied not only to text
ads but also to display ads. We also believe that industry should
continue working together to provide, for example, effective mechanisms
that empower consumers with the ability to opt out of behaviorally
targeted advertising.
Development of Technology to Empower Users
Products like Google Toolbar let a user choose to not have data
collected, and that choice persists even if all cookies are cleared and
until the user chooses to have data collected. Google also offers
features like Web History, which allows users to view and search all
search queries they have made on Google search while logged into
Google. Web History also lets users delete and thus disassociate from
their account information any searches that they conduct while they are
logged in. Users can also pause Web History altogether if they do not
want their searches to be associated with their account information--
and this choice persists until users choose to resume Web History. We
believe that more can be done by industry to ensure the persistence of
users' choices, and we look forward to exploring such tools with
industry and other stakeholders.
Conclusion
Chairman Inouye, Vice Chairman Stevens, and Members of the
Committee, thank you for the opportunity to testify today. I appreciate
the opportunity to explain the benefits of our advertising business to
consumers, advertisers, and publishers, and the chance to explain how
Google protects our users' privacy.
I look forward to answering any questions you might have about our
efforts, and Google looks forward to working with Members of the
Committee and others in the development of better privacy protections
for Internet users everywhere.
Thank you.
Senator Dorgan. Ms. Horvath, does that complete your
testimony?
Ms. Horvath. Yes. Thank you very much.
Senator Dorgan. Ms. Horvath, thank you very much.
Next we will hear from Mr. Robert Dykes, who is the
Chairman and CEO of NebuAd, Incorporated. Mr. Dykes, welcome.
You may proceed.
STATEMENT OF ROBERT R. DYKES, FOUNDER, CHAIRMAN, AND CEO,
NebuAd, INC.
Mr. Dykes. Thank you, Chairman Dorgan and Senator Vitter.
My name is Bob Dykes, CEO of NebuAd, Inc., a recent entrant
into the online advertising industry that partners with
Internet service providers, otherwise known as ISP's. I come
from a security background, serving for many years as Executive
Vice President of Symantec Corporation.
When we launched NebuAd several years ago, it was at a time
when many people had particularly heightened concerns about
data security. As part of its mission, NebuAd sought to address
these privacy and security concerns. As you will see, NebuAd
systems are designed so that no one, not even the government,
can determine the identity of our users.
Currently online advertising solutions and data collection
methods operate in many locations throughout the Internet
ecosystem, from users' computers to individual websites, to
networks of websites. The NebuAd service, in partnership with
ISP's, provides consumers with significant benefits serving
them with more relevant ads which they want, while ensuring
that they have robust privacy protections and control over
their online experience.
NebuAd's ad network also is designed to benefit two groups
that provide substantial benefit to the Internet: many smaller
websites and general use sites that have difficulty maintaining
free access to their content; the ISP's who need to upgrade
their infrastructure to provide increased bandwidth for
consumers who increasingly want access to Internet delivered
videos. NebuAd creates these benefits by using a select set of
a user's Internet activities to construct anonymous inferences
about likely interests which are then used to select and serve
the most relevant advertisements.
The NebuAd service is architected and its operations are
based on principles essential to strong privacy protection. We
provide users with prior robust notice about the service and
opportunity to express informed choice about whether to
participate both before the service takes effect and
persistently thereafter. We do not collect or use personally
identifiable information, or PII. We do not store raw data
linked to identifiable individuals, and we provide state-of-
the-art security for the limited amount of information we do
store. In other words, allegations by others that we do not
provide an opportunity to opt out in our robust notice to users
or that we collect entire web traffic of users are simply not
accurate.
To repeat, NebuAd provides robust notice and the
opportunity to express informed choice, and there is no
collection or use of any personally identifiable information or
even a significant portion of users' web traffic, nor any
information from password-protected sites, web mail, e-mail,
instant messages, or VoIP traffic.
We understand that to gain the public's trust, we need to
adopt strong privacy protections. Ours have been reviewed by
such entities as the Ponemon Institute, and we are engaging a
Big Four audit firm to conduct an audit to verify that we do
what we say we do.
This Committee has long been involved with the creation of
privacy statutes covering the cable and telecommunications
industries, as well as specific statutes addressing online
privacy for children, telemarketing, and spam. Yet even though
these and other privacy statutes have been developed one at a
time, there are common threads running through them all. When
more sensitive data is collected and when the collection and
disclosure of the data could harm or embarrass a consumer, more
rigorous disclosure and consent requirements tend to be
imposed. When raw data linked to an identifiable individual is
stored for longer periods, there is an emerging trend that more
rigorous disclosure, consent, and security requirements should
be imposed.
NebuAd supports the privacy paradigm which provides users
with consistent expectations and substantial protections. This
paradigm also is technology- and business-neutral, and it is
the basis upon which NebuAd built its technology and
operations. NebuAd urges the Committee to maintain both the
paradigm and the principle of technology and business
neutrality.
Thank you.
[The prepared statement of Mr. Dykes follows:]
Prepared Statement of Bob Dykes, CEO, NebuAd, Inc.
Chairman Inouye, Ranking Member Stevens, and Members of the
Committee, thank you for inviting me to appear today regarding the
privacy implications of online advertising. My name is Bob Dykes, CEO
of NebuAd, Inc., a recent entrant into the online advertising industry
that partners with Internet Service Providers (ISPs). I have spent
considerable time over the past year with Federal policymakers at the
Federal Trade Commission (FTC), Federal Communications Commission, and
in Congress--as well as with consumer and privacy advocates--discussing
NebuAd's technology, operations, and privacy protections and welcome
the opportunity to discuss all of this further with the Committee.
Introduction
Online advertising is a phenomenon of the Internet age. It permits
advertisers to provide more relevant messages to consumers and in turn
fuels the development of website publishers, both large and small. In
fact, advertising is the engine for the free Internet. The FTC has
found online advertising benefits consumers by enabling ``access to
newspapers and information around the world, provided free because it
is subsidized by online advertising; tailored ads that facilitate
comparison shopping for the specific products that consumers want; and,
potentially, a reduction in ads that are irrelevant to consumers'
interests and that may therefore be unwelcome.'' \1\
---------------------------------------------------------------------------
\1\ It is an axiom that advertising has more value when the
advertiser believes the user is more interested in the advertiser's
product. Such interest is not obvious when a user visits general-
purpose news and information sites, which are some of the very ones
noted by the FTC Staff as standing to benefit from online advertising.
Accordingly, the online advertising industry is constantly seeking
other ways to infer user interest and then bring that knowledge to bear
on the placement of ads on these sites. That is, behavioral advertising
drives value and supports those sites on the Internet that provide
society with great value.
---------------------------------------------------------------------------
Within this world of online advertising, NebuAd is a newcomer, just
entering among industry giants like Google, Yahoo!, Microsoft, Amazon,
and countless website publishers. That means we have a steep hill to
climb, but it also means we have great opportunities. We are able to
learn the lessons of the industry and construct state-of-the-art
technology that delivers ads that are more relevant to users while
providing them with robust and industry-leading privacy protections.
Indeed, as I will discuss, these privacy protections are built into our
technology and designed into our policies from the ground up.
Let me explain our privacy motivation more fully. I come from a
security background, serving for many years as Executive Vice President
of Symantec Corporation, a global leader in providing security
solutions for computers and computer networks. When we launched NebuAd
several years ago, it was at a time when many people had particularly
heightened concerns about data security. Hackers were piercing
firewalls, seeking to capture seemingly random strands of data to find
the identity of users. The government was ordering ISPs and other
network providers to turn over data on their users. As part of its
mission, NebuAd sought to address these privacy and security concerns.
The NebuAd service is architected and its operations are based on
principles essential to strong privacy protection:
Provide users with prior, robust notice and the opportunity
to express informed choice about whether to participate, both
before the service takes effect and persistently thereafter;
Do not collect or use personally-identifiable information
(``PII'');
Do not store raw data linked to identifiable individuals;
and
Provide state-of-the art security for any information
stored.
As a result, NebuAd's service is designed so that no one--not even
the government--can determine the identity of our users. That means our
service for ISP users, including the ad optimization and serving
system, does not collect or use any PII. In addition, NebuAd requires
its Internet service provider (``ISP'') partners to provide robust,
advance notice about our operations and our privacy protections to
their subscribers, who at any time can exercise their choice not to
participate. And, finally, we have located our servers in highly secure
data centers.
The NebuAd Technology and its Advertising Operations
Currently, online advertising solutions operate in many locations
throughout the Internet ecosystem--from users' computers to individual
websites to networks of websites. When an Internet user visits the
sites of web publishers, like Yahoo! or Amazon, these sites typically
collect information about the user's activities to target ads based on
that information. When an Internet user conducts a search, the search
company may collect information from the user's activity, which in turn
may be used to improve the relevance of the ads shown. And when a user
visits a website within an online advertising network, some of which
include thousands of sites, the visits help the network advertising
company categorize a user for targeted advertising. All of these
activities are well-entrenched in the Internet and, given the enormous
and growing use of the Internet, have proven to have mutual benefits
for users, publishers--large and small--advertisers, and ad networks.
NebuAd provides online advertising in partnership with ISPs. The
NebuAd advertising service has been architected to use only a select
set of a user's Internet activities (only a subset of HTTP traffic) to
construct anonymous inferences about the user's level of qualification
for a predefined set of market segment categories (``anonymous user
profiles''), which are then used to select and serve the most relevant
advertisements to that user. The NebuAd advertising service does not
collect or use any information from password-protected sites (e.g.,
HTTPS traffic), web mail, e-mail, instant messages, or VoIP traffic.
Using only non-PII, NebuAd constructs and continuously updates these
unique and anonymous user profiles.\2\
---------------------------------------------------------------------------
\2\ The anonymous user profiles do not contain any original raw
data, such as URLs navigated, but only consist of a set of numbers that
represent the anonymous inferences about the user's level of
qualification for a predefined set of market segment categories.
---------------------------------------------------------------------------
In the course of these business operations, NebuAd's ad
optimization and serving system does not collect PII or use information
deemed to be sensitive (e.g., information involving a user's financial,
sensitive health, or medical matters).\3\ In addition, NebuAd requires
its ISP partners to provide robust disclosure notices to users prior to
initiating any service and permits them to opt-out of having their data
collected and receiving targeted ads. Once a user opts-out, NebuAd
deletes that user's anonymous user profile and will ignore the user's
subsequent web navigation activity.\4\
---------------------------------------------------------------------------
\3\ NebuAd understands that the definition of ``sensitive''
information will evolve. We stated in our comments to the FTC on the
``Staff's Proposed Principles for the Self-Regulation of Behavioral
Advertising'' that we would adopt the Staff's definition of
``sensitive'' information, assuming it is not limitless. We also would
consider additional reasonable limitations proposed by other
stakeholders.
\4\ NebuAd has enhanced the industry-standard opt-out ``cookie''
based system with the use of proprietary techniques. This enables the
opt-out to be more persistent. NebuAd's entire enhanced opt-out system
is linked to individual computers and browsers, and it informs users of
this fact in assisting them in understanding the nature of their opt-
out choice.
---------------------------------------------------------------------------
Finally, NebuAd's ad optimization and serving system operates
similar to traditional ad networks. It makes standard use of cookies
for accepted ad serving purposes. It makes standard use of pixel tags
that operate only within the security framework of the browser to
invoke the placement of ad network cookies and that contain no uniquely
identifying number, subscriber identifier, or any other subscriber
information. In sum, NebuAd's code used for standard ad serving
purposes is both clean in its purpose and function.
The Privacy Paradigm in the United States and NebuAd's Privacy
Protections
In contrast to the European Community, where omnibus privacy law
covers all industries, in the United States, privacy statutes have been
developed in a largely sector-specific fashion. This Committee has long
been part of that trend, having overseen the creation of privacy
statutes generally covering the cable and telecommunications
industries, as well as specific statutes addressing online privacy for
children, telemarketing, and spam. Yet, even though these and other
privacy statutes have been developed one at a time, there are common
threads running through them:
When more sensitive data is collected, and when the
collection and disclosure of the data could harm or embarrass a
consumer, more rigorous disclosure and consent requirements
tend to be imposed.
When raw data linked to an identifiable individual is stored
for longer periods, there is an emerging trend that more
rigorous disclosure, consent, and security requirements should
be imposed.
NebuAd supports this privacy paradigm, which provides users with
consistent expectations and substantial protections. This paradigm also
is technology and business-neutral, and it is the basis upon which
NebuAd built its technology and operations. NebuAd urges the Committee
to maintain both the paradigm and the principle of technology and
business-neutrality.
In implementing this privacy paradigm, NebuAd not only relied on
the expertise of its own personnel, it turned to leading privacy
experts, including Fran Maier, Executive Director and President of
TRUSTe, the consumer privacy organization, Dr. Larry Ponemon of the
Ponemon Institute, and Alan Chapell of Chapell & Associates. These
experts provided important input into NebuAd's initial privacy program.
They were particularly stringent in recommending that NebuAd should not
collect PH or sensitive information and that it provide consumers with
robust notice and choice. NebuAd followed that guidance in developing
our privacy program.\5\
---------------------------------------------------------------------------
\5\ A just released survey of U.S. Internet users by TRUSTe showed
that 71 percent of online consumers are aware their web-surfing
information may be collected for the purpose of advertising and 91
percent wanted to have the tools to assure they could protect their
privacy. NebuAd has strived to provide users with this transparency by
educating users about its activities and their choices regarding
whether to participate in NebuAd's services.
---------------------------------------------------------------------------
The following are the key privacy protections upon which NebuAd has
architected into its technology and based its operations:
1. NebuAd's service does not collect or use PII from ISP
subscribers. The entire ad optimization and serving system does
not collect or use any PII, nor does it collect any information
from password-protected sites, web mail, e-mail, instant
messages, or VoIP traffic.
2. NebuAd stores only a set of numbers that represent the
user's level of qualification for a predefined set of market
segment categories (``anonymous user profiles''). NebuAd does
not store raw data such as URLs navigated or IP addresses
associated with an identifiable individual. Rather, the NebuAd
service constructs anonymous inferences about the user's level
of qualification for a predefined set of market segment
categories, and then discards the raw data that was used to
create or update a user's anonymous profile. This mechanism of
constructing anonymous inferences about the user's level of
qualification and not storing raw data provides a strong
additional layer of privacy protection that goes beyond the
standards used by many Internet companies today.
3. NebuAd's ISP Partners are required to provide robust, direct
notice in advance of launch of the service. The notice
discloses to the user that the ISP is working to ensure that
advertisements shown will be more relevant advertisements, that
to deliver these ads its partner creates anonymous profiles
based on part of the user's web-surfing behavior, which does
not include the collection of PII, and that the user may opt-
out of the service. For existing subscribers, the notice is
required to be delivered 30 days prior to the launch of the
service by postal mail, e-mail, or both.\6\ For new
subscribers, the notice is required to be placed clearly and
conspicuously in the new subscriber sign-up flow and outside
the privacy policy. All subscribers can opt-out at any time,
and ongoing disclosure and opportunity to opt-out is required
to be provided within the ISP's privacy policy.
---------------------------------------------------------------------------
\6\ NebuAd seeks to ensure that users are fully informed of its
activities and are given full opportunity to choose whether to
participate. To that end, we are developing enhanced notification
mechanisms.
4. NebuAd and its ISP partners offer users advance and on-going
choice of opting-out of the service. Users are provided with a
clear statement of what opt-out means and the way it operates.
Once the opt-out option is chosen, NebuAd honors that choice
and ignores the user's subsequent web surfing activity and thus
does not serve the user with behaviorally targeted ads.\7\
---------------------------------------------------------------------------
\7\ The user, of course, will continue to receive ads.
5. NebuAd's service only creates anonymous user profiles, which
contain no PII and no raw data, and its placement of ads is
completely anonymous. NebuAd uses proprietary algorithms and
techniques, including one-way encryption of data, so that no
one--not even NebuAd's engineers who designed the system--can
reverse-engineer an anonymous identifier, or the anonymous user
---------------------------------------------------------------------------
profile associated with it, to an identifiable individual.
6. NebuAd avoids any sensitive websites or product categories.
NebuAd does not track or serve ads based on visits related to
adult content, sensitive medical information, racial or ethnic
origins, religious beliefs or content of a sexual nature, and
does not have market segment categories for illegal products.
7. NebuAd does not permit either complexity of data or
narrowness of data to be reverse-engineered into PII. This
protection is accomplished because anonymous user profiles are
constructed by anonymous inferences about the consumer's level
of qualification for a predefined set of market segment
categories. Raw data is simply not stored as part of the
anonymous user profile. In addition, the NebuAd service does
not have narrowly-defined segments. Finally, the anonymous
profile identifier is the result of multiple encryptions, and
based on multiple data elements including the hashed IP
address.
8. There is no connection or link between the ISP's
registration data systems and NebuAd. That means that no user-
specific data is exchanged between NebuAd and ISP data systems.
This boundary is preserved further and inadvertent disclosure
is prevented because NebuAd immediately performs a one-way
encryption of the IP address and other anonymous user
identifiers used within the NebuAd system.
9. NebuAd installs no applications on users' computers, has no
access to users' hard drives, and has no access to secure
transactions. As such, NebuAd does not control a user's
computer or web-surfing activity in any way (e.g., by changing
computer settings or observing private or sensitive
information).
10. NebuAd's Data Centers are professionally operated and
secured. NebuAd's servers are located at secure sites with
state-of-the-art protections against any intrusion, electronic
or physical.
NebuAd is proud of these protections--all of which were adopted to
comply with both the spirit and letter of the government's privacy
paradigm--and, it continuously seeks to enhance them.
Conclusion
As I stated at the outset, I have spent years seeking to ensure
that users have robust and transparent privacy protections. In a very
real sense, NebuAd is the product of that work--It has adopted and
implemented state-of-the-art privacy protections, and, equally as
important, it has established a process to continuously improve on
them. The Internet is a highly dynamic environment, where new
technologies are constantly developed to address new challenges, and we
both want and need to take advantage of them. NebuAd takes its
responsibilities to its users very seriously. It looks forward to
continuing to work with government policymakers as they examine online
advertising and privacy issues.
Senator Dorgan. Mr. Dykes, thank you very much. We
appreciate your being here.
Next we will hear from Ms. Leslie Harris, who is the
President and Chief Executive Officer of the Center for
Democracy and Technology. Ms. Harris, you may proceed.
STATEMENT OF LESLIE HARRIS, PRESIDENT AND CEO, CENTER FOR
DEMOCRACY AND TECHNOLOGY
Ms. Harris. Thank you. Chairman Dorgan, Members of the
Committee, I thank you for the opportunity testify today.
I want to make three points and offer several
recommendations for the Committee.
First, while behavioral advertising is growing, consumers
are largely uncomfortable with the practice and they are ill-
equipped to take meaningful steps to protect their privacy. We
do recognize that advertising is an important engine of
Internet growth and that ad revenue supports a rich diversity
of online content. However, massive increases in data
processing and storage have allowed advertisers to track,
collect, and aggregate information about consumers' web
browsing across sites over time and to compile individual
profiles, and while each piece of consumer information in a
profile may itself not be personally identifiable, the
aggregation of this information into rich profiles means it may
be more readily tied to a person's identity.
All this is happening in an environment where more data is
being collected, retained for longer periods of time, often in
a form where it can be re-identified. Existing privacy
protections are outstripped by technology and there is a lack
of transparency about behavioral advertising practices and an
absence of meaningful controls to help consumers make informed
decisions about the use of their data.
When consumers do find out about behavioral advertising,
they are uncomfortable. In a recent study, 59 percent of the
respondents said they were not comfortable with online
companies using their browsing behavior to target advertising
to their interests even when they were told that advertising
supports free services.
Second, there is an emerging behavioral advertising model
that partners ISPs with ad networks, and this does add new
legal complexity and additional risks for privacy. While online
websites and networks can track what you see on their sites or
the sites associated with their networks, an ISP model may--and
I do say ``may''--provide access to everything you do online.
And that would include noncommercial sites that could reveal
political preferences, charitable, religious associations, and
the like.
Consumers simply do not have an expectation that their web
traffic is being intercepted by their ISP and shared with an
unknown third party for profiling. And in our view, the law
does not permit it. We read the Wiretap Act which prohibits
interception and disclosure of electronic communications,
including the content of Internet traffic, to require
unavoidable notice and affirmative express consent for ISP-
based behavioral targeting. And many state communications laws
require two-party consent, which further complicates the legal
landscape. And of course, the Cable Communications Policy Act
also prohibits the collection or disclosure of personal
information without prior consent.
The implementation that we have seen thus far completely
fails to acknowledge, let alone comply with, these laws. No one
has sought consent.
Finally, self-regulation is not a full answer. CDT has
always been supportive of self-regulation, but the Network
Advertising Initiative that was launched 8 years ago has been
largely a failure. The model fell short when it was announced.
It has failed to evolve over time. And only now when the FTC
has turned its attention to the issues has the Initiative
proposed modest improvements.
We acknowledge that there are a number of individual
companies that have worked hard to educate and improve their
users' privacy. We are prepared to continue to work with them
and the NAI to continue to improve privacy protection.
Self-regulation was never expected to be a full solution.
And when the NAI was created, the FTC at that point noted that
backstop legislation would still be required to ensure
consumers' privacy is protected.
We have made a number of recommendations for Committee
action in our written testimony. Let me just briefly highlight
a few.
First, Senator Dorgan, we agree with you that more hearings
are necessary on the ISP question and on other questions like
sensitive information and secondary use.
Second, we really urge this Committee to set a goal of
enacting general privacy legislation in the next year based on
well-established fair information practices. We have been
advocating for this for years. We think it is time to act.
Third, we do think the FTC should issue its guidelines. We
think they need to be enforceable whether they are under
current authority or under targeted legislation. We think the
Committee should make that clear to the FTC.
And finally, we think Congress should encourage the FTC to
investigate how technological solutions, including perhaps a
``Do Not Track'' regime--we were part of a group that proposed
that--can give consumers better control over their online
information.
We do think Congress has a critical role to play in
ensuring privacy protection in this increasingly complex online
advertising environment, and I look forward to answering your
questions.
[The prepared statement of Ms. Harris follows:]
Prepared Statement of Leslie Harris, President/CEO,
Center for Democracy and Technology
Chairman Inouye and Members of the Committee:
On behalf of the Center for Democracy and Technology (``CDT''), I
thank you for the opportunity to testify today. We applaud the
Committee's leadership in examining the privacy impact of new online
advertising models.
I. Summary
CDT recognizes that advertising is an important engine of Internet
growth. Consumers benefit from a rich diversity of content, services
and applications that are provided without charge and supported by
advertising revenue. However, as sophisticated new behavioral
advertising models are deployed, it is vital that consumer privacy be
protected. Massive increases in data processing and storage
capabilities have allowed advertisers to track, collect and aggregate
information about consumers' web browsing activities, compiling
individual profiles used to match advertisements to consumers'
interests. All of this is happening in the context of an online
environment where more data is collected--and retained for longer
periods--than ever before and existing privacy protections have been
far outpaced by technological innovation.
Behavioral advertising represents a small but rapidly growing part
of the online advertising market. Market research firm eMarketer
reported last year that spending on behaviorally targeted online
advertising is expected to reach $1 billion this year and to quadruple
by 2011.\1\ The recent spate of acquisitions of the online advertising
industry's largest players by major Internet companies is powerful
evidence that the online advertising marketplace is headed toward more
data aggregation tied to a single profile--and one that may be more
readily tied to a person's identity.\2\ And while we have yet to see
evidence that this new advertising model will reap the promised
rewards, it is already migrating from individual websites to the
infrastructure of the Internet itself: In the last year, Internet
Service Providers (``ISPs'') have begun to form partnerships with ad
networks to mine information from individual web data streams for
behavioral advertising. Ad networks that partner with ISPs could
potentially collect and record every aspect of a consumer's web
browsing, including every web page visited, the content of those pages,
how long each page is viewed, and what links are clicked. E-mails,
chats, file transfers and many other kinds of data could all be
collected and recorded.
---------------------------------------------------------------------------
\1\ ``Behavioral Advertising on Target . . . to Explode Online,''
eMarketer (Jun. 2007), http://www.emarketer.com/
Article.aspx?id=1004989.
\2\ No fewer than five major mergers and acquisitions have been
completed in the last 18 months: Google purchased online advertising
company DoubleClick, Inc.; WPP Group, a large ad agency, acquired the
online ad company 24/7 Real Media; Yahoo! acquired ad firm RightMedia;
Microsoft acquired online ad service provider aQuantive; AOL purchased
Tacoda, a pioneering firm in the area of behavioral advertising.
---------------------------------------------------------------------------
The ISP model raises particularly serious questions. Thus far,
implementations appear to defy reasonable consumer expectations, could
interfere with Internet functionality, and may violate communications
privacy laws.
Notwithstanding the recent growth of behavioral advertising, most
Internet users today do not know that their browsing information may be
tracked, aggregated and sold. After almost a decade of self-regulation,
there is still a profound lack of transparency associated with these
practices and an absence of meaningful consumer controls.
There are several efforts underway to respond to the new online
advertising environment. First, the Federal Trade Commission staff
recently released a draft of proposed principles for self-regulation,
which represent a solid step forward. However, it is not clear whether
the FTC will formally adopt the principles or put its enforcement power
behind them.
The Network Advertising Initiative (``NAI'') is also in the process
of revising its guidelines. This is a welcome but long-overdue
development. Unfortunately, self-regulation has not worked to date and,
even if strengthened, will never by itself fully protect consumers'
privacy interests.
Congress needs to take a comprehensive look at the current and
emerging practices associated with behavioral advertising and the risks
those practices pose to consumer privacy and control. We recommend that
Congress take the following steps to address the significant privacy
concerns raised by behavioral advertising:
The Committee should hold a series of hearings to examine
specific aspects of behavioral advertising, in particular the
growing involvement of ISPs, the use of sensitive information,
and secondary uses of behavioral profiles.
The Committee should set a goal of enacting in the next year
a simple, flexible baseline consumer privacy law that would
protect consumers from inappropriate collection and misuse of
their personal information, both online and offline.
The Committee should strongly urge the Federal Trade
Commission to exercise its full enforcement authority over
online advertising practices.
Congress should examine and strengthen existing
communications privacy laws to cover new services, technologies
and business models with consistent rules. The Electronic
Communications Privacy Act (``ECPA'') is decades old, and its
application in today's online world is often unclear.
Congress should encourage the FTC to investigate how
technology can be harnessed to give consumers better control
over their online information. Simple tools that put consumers
in controls of their information, such as a ``Do Not Track''
list, deserve consideration.
II. Understanding Online Advertising Practices
Commercial websites that supply content to consumers free of charge
are often supported by online advertising. These sites--known as
``publishers'' in the advertising world--make available certain
portions of space on their pages to display ads. That space is sold to
advertisers, ad agencies, or online ad intermediaries that find and
place advertisements into the space. These intermediaries may also make
arrangements to collect information about user visits to the publisher
pages. Since very few publishers supply their own advertising, it is
common that when a consumer visits a publisher site, the consumer's
computer also connects to one or more advertisers, ad agencies, or ad
intermediaries to send data about the consumer's visit to the site and
receive the advertising on the site.
One type of ad intermediary is known as an ``advertising network.''
At their most basic level, ad networks contract with many different
publishers on one side and many different advertisers on the other.
Armed with a pool of space in which to display ads on publisher sites,
and a pool of ads to display, ad networks are in the business of
matching up the two by using the data they collect about consumers'
site visits.
A. Contextual Advertising
There are many different ways for an ad network to determine which
advertisement should be placed in which space. The two most often
discussed are ``contextual'' advertising and ``behavioral''
advertising. Contextual advertising, which is often used to generate
ads alongside search results, matches advertisements to the content of
the page that a consumer is currently viewing--a consumer who visits a
sports site may see advertisements for golf clubs or baseball tickets
on that site.
The privacy risks associated with contextual advertising vary. If
the practice is transparent to the user and data collection and
retention is minimal, the practice poses little risk. By contrast,
privacy concerns are heightened if the user data is retained in an
identifiable or pseudonymous form (i.e., linked to a user identifier)
for long periods of time even if it is not immediately used to create
advertising profiles.
B. Behavioral Advertising
By contrast, behavioral advertising matches advertisements to the
interests of the consumer as determined over time. If a consumer visits
several different travel sites before viewing a news site, he or she
might see a behaviorally targeted travel advertisement displayed on the
news page, even if the news page contains no travel content. A
traditional behavioral ad network builds up profiles of individual
consumers by tracking their activities on publisher sites in the
network (although this model is evolving, as we discuss below). When
the consumer visits a site where the ad network has purchased ad space,
the ad network collects data about that visit and serves an
advertisement based on the consumer's profile. Diagrams illustrating
this process are included in Appendix A.
Consumers' behavioral advertising profiles may incorporate many
different kinds of data that are in and of themselves not personally
identifiable. Many networks avoid linking profiles to what has
traditionally been considered ``personally identifiable information''
(``PII''): names, addresses, telephone numbers, e-mail addresses, and
other identifiers. But as the comprehensiveness of consumer advertising
profiles increases, the ability of marketers and others to link
specific individuals to profiles is also growing. In 2006, for example,
AOL released 3 months' worth of search queries generated by half a
million users; in the interest of preserving users' anonymity, AOL
replaced individuals' screen names with numbers. Based solely on search
terms associated with one number, reporters at The New York Times were
able to pinpoint the identity of the user who generated them.\3\ The
risk of supposedly non-personally identifying data being used to
identify individuals has spurred several ad networks to take extra
steps to de-identify or remove personal information from their data
storage.\4\
---------------------------------------------------------------------------
\3\ Michael Barbaro and Tom Zeller, Jr., ``A Face Is Exposed for
AOL Searcher No. 4417749,'' The New York Times (Aug. 2006), http://
www.nytimes.com/2006/08/09/technology/ 09aol.
html?_r=1&ex=1312776000&adxnnl=1&oref=slogin&adxnnlx=1215021816
2Dj7kbrLxHU1hCdcMyNqHEbA.
\4\ See, e.g., Microsoft, Privacy Protections in Microsoft's Ad
Serving System and the Process of ``De-identification'' (Oct. 2007),
http://download.microsoft.com/download/3/1/d/31df6942-ed99
-4024-a0e0-594b9d27a31a/
Privacy%20Protections%20in%20Microsoft%27s%20Ad%20serving%20
system%20 and%20the%20Process%20of%20De-Identification.pdf.
---------------------------------------------------------------------------
Profiles may also be intentionally tied to PII. For example, data
collected online by a merchant or by a service provider may permit an
advertising profile to be tied to an individual's e-mail account.
Offline data may also be merged with online profiles. For years, data
service companies have maintained profiles about consumers based on
information gleaned from public sources such as property and motor
vehicle records, as well as records from sources like catalog sales and
magazine subscriptions. These data companies are now also entering the
online advertising business, potentially allowing the linking of online
and offline profiles.\5\
---------------------------------------------------------------------------
\5\ Acxiom runs Relevance-X, an online ad network. Last year
Experian acquired the online data analysis company Hitwise. See Acxiom,
Acxiom: Relevance-X (last visited Jul. 2008),
http://www.acxiom.com/Relevance-X; Experian, ``Acquisition of Hitwise''
(Apr. 2007), http://www.experiangroup.com/corporate/ news/releases/
2007/2007-904-17b/.
---------------------------------------------------------------------------
C. The Evolution of Behavioral Advertising--More Data, More Data
Sources
As noted above, recent market consolidation facilitates more
comprehensive data collection. Companies that run consumers' favorite
web-based services--web search, web mail, maps, calendars, office
applications, and social networks--have all purchased behavioral
advertising networks within the last year. In the past, major Internet
companies could gather information about how an individual used its
services and applications such as search, but did not have direct
access to information about the user's other web browsing habits. With
the acquisition of behavioral advertising networks, these companies
could potentially marry the rich data about an individual's use of one
site with a broad view of his or her activities across the web. The
concerns about this aggregation of consumer data are heightened because
many online companies retain data for months or years on end in
identifiable or pseudonymous form, creating a host of privacy risks.
Finally, ad networks are now turning to the most comprehensive and
concentrated source of information about Internet use: the individual
web data streams that flow through ISPs.\6\ In this emerging model, the
ISP intercepts or allows an ad network to intercept the content of each
individual's web data stream. The ad network then uses this traffic
data for behavioral advertising, serving targeted ads to the ISP's
customers on publisher sites as the customers surf the web. We address
the unique issues posed by this advertising model in detail below.
---------------------------------------------------------------------------
\6\ See, e.g., Peter Whoriskey, ``Every Click You Make,'' The
Washington Post (Apr. 2008), http://www.washingtonpost.com/wp-dyn/
content/article/2008/04/03/AR2008040304052.html? nav=hcmodule; Saul
Hansell, ``I.S.P. Tracking: The Mother of All Privacy Battles,'' The
New York Times: Bits Blog (Mar. 2008) at http://bits.blogs.nytimes.com/
2008/03/20/isp-tracking-the-mother-of-all-privacy-battles/?scp =1-
b&sq=the+mother+of+all+privacy+battles&st=nyt.
---------------------------------------------------------------------------
III. The Privacy Risks of Behavioral Advertising
Behavioral advertising poses a growing risk to consumer privacy;
consumers are largely unaware of the practice and are thus ill equipped
to take protective action. They have no expectation that their browsing
information may be tracked and sold, and they are rarely provided
sufficient information about the practices of advertisers or others in
the advertising value chain to gauge the privacy risks and make
meaningful decisions about whether and how their information may be
used. In a recently released Harris Interactive/Alan F. Westin study,
59 percent of respondents said they were not comfortable with online
companies using their browsing behavior to tailor ads and content to
their interests even when they were told that such advertising supports
free services.\7\ A recent TRUSTe survey produced similar results.\8\
It is highly unlikely that these respondents understood that this type
of ad targeting is already taking place online every day.
---------------------------------------------------------------------------
\7\ Alan F. Westin, How Online Users Feel About Behavioral
Marketing and How Adoption of Privacy and Security Policies Could
Affect Their Feelings (Mar. 2008).
\8\ TRUSTe, ``TRUSTe Report Reveals Consumer Awareness and
Attitudes About Behavioral Targeting'' (Mar. 2008), http://
www.marketwire.com/mw/release.do?id=837437&sourceType=1 (``71 percent
of online consumers are aware that their browsing information may be
collected by a third party for advertising purposes. . . . 57 percent
of respondents say they are not comfortable with advertisers using that
browsing history to serve relevant ads, even when that information
cannot be tied to their names or any other personal information.'').
---------------------------------------------------------------------------
In most cases, data collection for behavioral advertising operates
on an opt-out basis. Opt-out mechanisms for online advertising are
often buried in fine print, difficult to understand, hard to execute
and technically inadequate. Only the most sophisticated and technically
savvy consumers are likely to be able to successfully negotiate such
opt-out processes. Moreover, in most cases, opt-out mechanisms offered
for behavioral advertising only opt the user out of receiving targeted
ads, but do not opt the user out of data collection about his or her
Internet usage.
For behavioral advertising to operate in a truly privacy-protective
way, data collection needs to be limited and data retention limits
should be tied to the original purposes for collecting the data.
Consumers need to be informed about what data is being collected about
their Internet activities, how the information will be used, whether
the information will be shared with others, and what measures are being
taken to ensure that any transfer of data remains secure. They should
be presented with this information in a manner that supports informed
choice over their information and that choice should be honored
persistently over time. Consumers must also have opportunities for
legal redress for misuse of the data. As a recent D.C. District Court
opinion established, data leakage and the concern for potential abuses
of that data are recognizable harms standing alone, without any need to
show misuse of the data.\9\ Consumers do not need to become victims of
identity theft to suffer from an invasion of privacy.
---------------------------------------------------------------------------
\9\ Am. Fed'n of Gov't Employees v. Hawley, D.D.C., No. 07-00855,
3/31/08 (ruling, inter alia, that concerns about identity theft,
embarrassment, inconvenience, and damage to financial suitability
requirements after an apparent data breach constituted a recognizable
``adverse effect'' under the Privacy Act, 5 U.S.C. 552(a) (citing
Kreiger v. Dep't of Justice, 529 F.Supp.2d 29, 53 (D.D.C. 2008)).
---------------------------------------------------------------------------
There is also a risk that profiles for behavioral advertising may
be used for purposes other than advertising. For example, ad networks
that focus on ``re-targeting'' ads may already be using profiles to
help marketers engage in differential pricing.\10\ Behavioral profiles,
particularly those that can be tied to an individual, may also be a
tempting source of information in making decisions about credit,
insurance, and employment. While the lack of transparency makes it
almost impossible to know whether behavioral profiles are being used
for other purposes, the lack of enforceable rules around the collection
and use of most personal information leaves the door wide open for a
myriad of secondary uses.
---------------------------------------------------------------------------
\10\ See Louise Story, ``Online Pitches Made Just For You,'' The
New York Times (Mar. 2008), http://www.nytimes.com/2008/03/06/business/
media/06adco.html.
---------------------------------------------------------------------------
Finally, because the legal standards for government access to
personal information held by third parties are extraordinarily low,
these comprehensive consumer profiles are available to government
officials by mere subpoena, without notice to the individual or an
opportunity for the individual to object.\11\
---------------------------------------------------------------------------
\11\ See Center for Democracy and Technology, Digital Search &
Seizure: Updating Privacy Protections to Keep Pace with Technology
(2006), http://www.cdt.org/publications/digital-search-and-seizure.pdf
at 7-9; Deirdre K. Mulligan, ``Reasonable Expectations in Electronic
Communications: A Critical Perspective on the Electronic Communications
Privacy Act,'' 72 Geo. Wash. L. Rev. 1557 (Aug. 2004); Daniel J.
Solove, ``Digital Dossiers and the Dissipation of Fourth Amendment
Privacy,'' 75 S. Cal. L. Rev. 1083, 1135 (2002).
---------------------------------------------------------------------------
IV. The Use of Sensitive Information for Behavioral Advertising
The concerns about behavioral advertising practices are heightened
because of the increasingly sensitive nature of the information that
consumers are providing online in order to take advantage of new
services and applications. Two data types of particular concern are
health information and location information.
A. Personal Health Information--Increasingly Available Online
Personal health data is migrating online through an ever-expanding
array of health information and search sites, online support groups,
and personal health record sites. Federal privacy rules under the
Health Information Portability and Accountability Act (``HIPAA'') do
not cover personal health information once it moves online and out of
the control of HIPAA-covered entities. Once it is posted online, it may
have no more legal protection than any other piece of consumer
information. In addition, information provided by consumers that is not
part of a ``medical record''--such as search terms--may nevertheless
reveal highly sensitive information. We do not know the full extent to
which personal health data is being collected for behavioral
advertising. We do know that the limits placed on its collection by the
industry are inadequate and that there is an urgent need to develop a
definition for personal health information in the Internet context that
is robust enough to protect privacy.
B. Location Information--Not Always Protected By Current Law
As technologies converge and Internet services are provided over
cellular phones and other mobile devices, the ability to physically
locate consumers is spurring location-based advertising, targeted to
where a user is at any given moment. Plans to incorporate location
information into behavioral advertising are still in development.
Although laws exist to protect location information collected by
telecommunications carriers, applications providers are increasingly
offering location-based services that fall completely out of that legal
framework. Standards for government access to location information are
also unclear, even as law enforcement has shown a greater interest in
such information.\12\
---------------------------------------------------------------------------
\12\ See Center for Democracy and Technology, Digital Search &
Seizure: Updating Privacy Protections to Keep Pace with Technology
(2006), http://www.cdt.org/publications/digital-search-and-seizure.pdf
at 23-29.
---------------------------------------------------------------------------
V. The Emerging Use of ISP Data for Behavioral Advertising
The use of ISP data for behavioral advertising is one area that
requires close scrutiny from lawmakers. The interception and sharing of
Internet traffic content for behavioral advertising defies reasonable
user expectations, can be disruptive to Internet and Web functionality,
and may run afoul of communications privacy laws.
A. How ISP Data is Used for Behavioral Advertising
In this new model, an ad network strikes a deal with an ISP that
allows the network to receive the contents of the individual web
traffic streams of each of the ISP's customers. The ad network analyzes
the content of the traffic in order to create a record of the
individual's online behaviors and interests. As customers of the ISP
surf the Web and visit sites where the ad network has purchased ad
space, they see advertisements targeted based on their previous
Internet behavior. While the model as it exists today involves an ISP
contracting with a third party that operates such an ad network, it
would also be possible for ISPs to do the traffic content inspection,
categorization, and advertising delivery themselves.
B. Privacy Implications of the Use of ISP Data for Behavioral
Advertising
The privacy implications of behavioral advertising at large are
amplified in this ISP model. Ad networks that partner with ISPs may
potentially gain access to all or substantially all of an individual's
Web traffic as it traverses the ISP's infrastructure, including traffic
to all political, religious, and other non-commercial sites. While
traditional ad networks may be large, few if any provide the
opportunity to collect information about an individual's online
activities as comprehensively as in the ISP model, particularly with
respect to activities involving non-commercial content. And although
these ad networks currently inspect predominantly Web traffic, ISPs
carry e-mails, chats, file transfers and many other kinds of data that
they could decide to pass on to behavioral ad networks in the future.
Moreover, the use of Internet traffic content for behavioral
advertising defies user expectations about what happens when they surf
the Web and communicate online. Absent unmistakable notice, consumers
simply do not expect their ISP or its partners to be looking into the
content of their Internet communications. Finding out that there is a
middleman lurking between consumers and the websites they visit would
come as a unwelcome surprise to most Internet users. ISPs are a
critical part of the chain of trust that undergirds the Internet.
Giving an unknown third party broad access to all or most consumer
communications may undermine that trust.
C. Current Implementations May Interfere With Normal Internet Use
Despite these concerns, several ad network companies are moving
forward with plans to use ISP data for behavioral advertising. The two
most prominent ad networks engaged in this practice are NebuAd in the
United States and Phorm in the UK. Charter Communications, a cable
broadband ISP, recently announced--and then delayed--a plan to conduct
trials of the NebuAd behavioral advertising technology.\13\ Several
other ISPs, such as Wide Open West (WOW!), CenturyTel, Embarq and
Knology also announced plans with NebuAd to trial or deploy its
behavioral advertising technology. Although a number of these ISPs have
put their plans on hold in the wake of a firestorm of criticism, NebuAd
continues to work with U.S. ISPs and seek new ISP partners. Phorm,
which originally announced deals with three of the UK's largest ISPs
and has sought partnerships with U.S. ISPs, is also now encountering
hesitation from some of its partners.\14\
---------------------------------------------------------------------------
\13\ Saul Hansell, ``Charter Suspends Plan to Sell Customer Data to
Advertisers,'' The New York Times: Bits Blog (Jun. 2008), http://
bits.blogs.nytimes.com/2008/06/24/charter-suspends-plan-to-sell-
customer-data-to-advertisers/?scp=3-b&sq=charter+nebuad&st=nyt.
\14\ Chris Williams, ``CPW builds wall between customers and
Phorm,'' The Register (Mar. 2008), http://www.theregister.co.uk/2008/
03/11/ phorm_shares_plummet/.
---------------------------------------------------------------------------
Independent analyses of both companies' systems have revealed that
by virtue of their ability to intercept Internet traffic in the middle
of the network--and based on their desire to track individual Internet
users--they engage in an array of practices that are inconsistent with
the usual flow of Internet traffic. NebuAd reportedly injects computer
code into Web traffic streams that causes numerous cookies to be placed
on users' computers for behavioral tracking, none of which are related
to or sanctioned by the websites the users visit.\15\ When a user
navigates to a particular website, Phorm reportedly pretends to be that
website so that it can plant a behavioral tracking cookie linked to
that site on the user's computer.\16\ In addition to the privacy
implications of tracking all of an individual's Web activities, this
kind of conduct has the potential to create serious security
vulnerabilities in the network,\17\ hamper the speed of users' Internet
connections, and interfere with ordinary Web functionality. At a time
when many different kinds of companies are working to build a trusted
computing platform for the Internet, having ISPs work with partners
whose practices undermine trust raises future cyber-security concerns.
---------------------------------------------------------------------------
\15\ Robert M. Topolski, NebuAd and Partner ISPs: Wiretapping,
Forgery and Browser Hijacking, Free Press and Public Knowledge (Jun.
2008), http://www.publicknowledge.org/ pdf/nebuad-report-20080618.pdf.
\16\ Richard Clayton, The Phorm ``Webwise'' System (May 2008),
http://www.cl.cam.ac.uk/rnc1/080518-phorm.pdf.
\17\ These types of behaviors have much in common with well-
understood online security threats, and parts of the Internet security
community are already investigating how to respond. See Anti-Spyware
Coalition, ``Anti-Spyware Coalition Aims to Address Behavioral
Targeting'' (Apr. 2008), http://antispywarecoalition.org/newsroom/
20080425press.htm.
---------------------------------------------------------------------------
D. Current Implementations May Violate Federal Law
Depending on how this advertising model is implemented, it may also
run afoul of existing communications privacy laws. The Federal Wiretap
Act, as amended by the Electronic Communications Privacy Act
(``ECPA''), prohibits the interception and disclosure of electronic
communications--including Internet traffic content--without
consent.\18\ Although exceptions to this rule permit interception and
disclosure without consent, we seriously doubt that any of them apply
to the interception or disclosure of Internet traffic content for
behavioral advertising purposes. Accordingly, we believe that the
Wiretap Act requires unavoidable notice and affirmative opt-in consent
before Internet traffic content may be used from ISPs for behavioral
advertising purposes. Certain state laws may take this one step
further, requiring consent from both parties to the communication: the
consumer and the website he or she is visiting. A detailed CDT legal
memorandum on the application of the Wiretap Act, ECPA and relevant
state wiretap laws to the use of ISP data for behavioral advertising is
attached as Appendix B.
---------------------------------------------------------------------------
\18\ 18 U.S.C. 2511.
---------------------------------------------------------------------------
As several Members of Congress have noted, the Cable Communications
Policy Act also applies here.\19\ The law prohibits cable operators
from collecting or disclosing personally identifiable information
without prior consent \20\ While the term ``personally identifiable
information'' in the law is defined by what it does not include--``any
record of aggregate data which does not identify particular persons''
\21\--we doubt that a user's entire Web traffic stream, unique to that
individual, often containing both PII and non-PII, would be considered
aggregate data as that term is commonly understood.
---------------------------------------------------------------------------
\19\ House Representative Edward Markey and House Representative
Joe Barton, Letter to Charter Communications CEO in Regards to the
Charter-NebuAd Data Collection Scheme (May 2008) http://
markey.house.gov/docs/telecomm/ letter_charter_comm_privacy.pdf. A 1992
amendment adding the phrase ``other services'' to the Cable Act's
privacy provision made it clear that the law covers Internet services
provided by cable operators.
\20\ 47 U.S.C. 551(b)-(c).
\21\ Id. 551(a)(2)(A).
---------------------------------------------------------------------------
We do not believe that it is possible to shoehorn the collection
and disclosure of a subscriber's entire browsing history for
advertising purposes into the statute's exception for collection or
disclosure of information that is necessary to render service.\22\
Thus, we conclude that cable-based ISPs that wish to disclose customer
information to advertising networks would also have to meet the consent
requirements of the Cable Communications Policy Act.
---------------------------------------------------------------------------
\22\ Id. 551(a)(2)(B).
---------------------------------------------------------------------------
The ISP models that have been deployed thus far have failed to
obtain affirmative, express opt-in consent required by law. Several
small U.S. ISPs, for example, have failed to meet this threshold
requirement, burying vague information about their deals with NebuAd in
the ISPs' terms of service.\23\ Charter Communications, the largest
U.S. ISP that had planned to partner with NebuAd, notified its
subscribers that they would be receiving more relevant ads, but did not
explain its plans to intercept subscribers' traffic data, and did not
provide a way for subscribers to give or withhold consent. Charter has
since suspended its plans.
---------------------------------------------------------------------------
\23\ See Mike Masnick, ``Where's The Line Between Personalized
Advertising And Creeping People Out?,'' TechDirt (Mar. 2008), http://
www.techdirt.com/articles/ 20080311/121305499.shtml; Peter Whoriskey,
``Every Click You Make,'' The Washington Post (Apr. 2008), http://
www.washingtonpost.com/wp-dyn/content/article/2008/04/03/
AR2008040304052.html?nav=hc
module.
---------------------------------------------------------------------------
Designing a robust opt-in consent system for ISP-based behavioral
advertising presents a formidable challenge. We are less than sanguine
that such a system can be easily designed, particularly since it must
not only provide a way for consumers to give affirmative consent, but
it must also provide a method for them to revoke that consent. The
burden is on those who wish to move forward with the model to
demonstrate that an express notice and consent regime can work in this
context.
VI. The Limits of Self-Regulation
For almost a decade, the primary privacy framework for the
behavioral advertising industry has been provided by the Network
Advertising Initiative, a self-regulatory group of online advertising
networks formed in response to pressure from the Federal Trade
Commission and consumer advocates in the wake of privacy concerns over
the merger of ad network DoubleClick and Abacus, an offline data
broker. NAI members agree to provide consumers with notice and, at
minimum, a method to opt out of behavioral advertising. They further
pledged to use information collected only for marketing purposes. While
at the time of their release CDT welcomed the NAI principles as an
important first step, we also noted then that there were flaws in the
approach that needed to be addressed and that self-regulation was not a
complete solution. The FTC agreed, concluding in its July 2000 report
to Congress that ``backstop legislation addressing online profiling is
still required to fully ensure that consumers' privacy is protected
online.'' \24\ That remains true today.
---------------------------------------------------------------------------
\24\ Federal Trade Commission, Online Profiling: A Report to
Congress (Jul. 2000), http://www.ftc.gov/os/2000/07/
onlineprofiling.htm.
---------------------------------------------------------------------------
Eight years after the creation of the principles, few consumers are
aware of behavioral advertising and fewer still have been able to
successfully navigate the confusing and complex opt-out process.\25\
Although individual NAI companies have launched their own consumer
awareness initiatives, more work remains to be done.\26\ For those
consumers who successfully opt out, the NAI's reliance on flawed opt-
out cookies means that user preferences are often not persistently
honored.
---------------------------------------------------------------------------
\25\ The drawbacks of opt-out cookies have been well documented:
they are confusing for the majority of consumers who do not understand
the technology and counter-intuitive to those who are accustomed to
deleting their cookies to protect their privacy. Cookies are
susceptible to accidental deletion and file corruption. While the NAI
is in the process of updating the principles, it has not proposed
changes to the opt-out regime. See Center for Democracy and Technology,
Applying the FTC's Spyware Principles to Behavioral Advertising:
Comments of the Center for Democracy and Technology in regards to the
FTC Town Hall, ``Ehavioral Advertising: Tracking, Targeting, and
Technology'' (Oct. 2007), http://www.cdt.org/privacy/
20071019CDTcomments
.pdf at 8.
\26\ See, e.g., AOL, Mr. Penguin (last visited Jul. 2008), http://
corp.aol.com/o/mr-penguin/; Yahoo!, Customized Advertising (last
visited Jul. 2008), http://info.yahoo.com/relevantads/; Google, The
Google Privacy Channel (last visited Jul. 2008), http://youtube.com/
user/googleprivacy.
---------------------------------------------------------------------------
In addition, the NAI's guidelines for the use of sensitive
information have never been adequate to guard consumer privacy. Until
recently, the definition was limited to a narrowly defined set of PII.
While the definition is being revised, it still falls far short of what
is needed to address the increasingly sensitive nature of consumer
information online.\27\
---------------------------------------------------------------------------
\27\ Center for Democracy and Technology, Comments Regarding the
NAI Principles 2008: The Network Advertising Initiative's Self-
Regulatory Code of Conduct for Online Behavioral Advertising (June
2008), http://www.cdt.org/privacy/20080612_NAI_comments.pdf at 6-9.
---------------------------------------------------------------------------
Finally, the NAI principles only apply to companies that
voluntarily join the Initiative. The NAI has no way to force companies
to join; the current membership is missing numerous behavioral
advertising firms, including some key industry players. In addition,
measures to ensure compliance and transparency have withered on the
vine.\28\ The original NAI principles provided for independent audits
and enforcement against noncompliant members, but the audit results
were never made public, and reporting on compliance with the principles
has been inconsistent.\29\
---------------------------------------------------------------------------
\28\ CDT testing has revealed that only a tiny fraction of
companies that collect data that could be used for behavioral
advertising are NAI members. See Center for Democracy and Technology,
Statement of The Center for Democracy and Technology before The
Antitrust, Competition Policy and Consumer Rights Subcommittee of the
Senate Committee on the Judiciary on ``An Examination of the Google-
DoubleClick Merger and the Online Advertising Industry: What Are the
Risks for Competition and Privacy?'' (Sept. 2007), http://www.cdt.org/
privacy/20070927committee-statement.pdf.
\29\ See Pam Dixon, The Network Advertising Initiative: Failing at
Consumer Protection and at Self-Regulation (Nov. 2007), http://
www.worldprivacyforum.org/pdf/WPF_NAI_report_
Nov2_2007fs.pdf at 16-17.
---------------------------------------------------------------------------
For all these reasons, while we encourage more robust self-
regulatory efforts, we continue to have doubts about the effectiveness
of the self-regulatory framework. As online advertising becomes
increasingly complex and data collection becomes more pervasive,
Congress and the FTC must step in to ensure that consumer interests are
fully protected.
VII. The Role of Congress
Congress should take action to address the significant privacy
concerns raised by behavioral advertising:
As a first step, we urge the Committee to hold a series of
hearings to examine specific aspects of behavioral advertising.
In particular, we believe that further investigation of new
models of behavioral advertising using ISP data is warranted,
and that the Committee should explore how current laws such as
ECPA, the Wiretap Act and the Cable Communications Policy Act
apply. Secondary uses of behavioral advertising profiles for
purposes other than marketing also deserve additional
investigation and scrutiny, as does the use of sensitive
information.
This Committee should set a goal of enacting in the next
year general privacy legislation covering both the online and
offline worlds. CDT has long argued for simple, flexible
baseline consumer privacy legislation that would protect
consumers from inappropriate collection and misuse of their
personal information while enabling legitimate business use to
promote economic and social value. In principle, such
legislation would codify the fundamentals of fair information
practices, requiring transparency and notice of data collection
practices, providing consumers with meaningful choice regarding
the use and disclosure of that information, allowing consumers
reasonable access to personal information they have provided,
providing remedies for misuse or unauthorized access, and
setting standards to limit data collection and ensure data
security.
The Federal Trade Commission has played a helpful role in
consumer education efforts around behavioral advertising. But
it also must exercise its authority under its deception and
unfairness jurisdiction to issue enforceable guidelines for
behavioral advertising. We ask the Committee to strongly urge
the Commission to exercise the full measure of its enforcement
authority over online advertising practices.
Congress should also examine and strengthen existing
communications privacy laws to cover new services, technologies
and business models with consistent rules. ECPA was passed more
than 20 years ago, long before there was a World Wide Web and
the Internet became integrated into Americans' daily lives. The
application of the law to common online activities including
Web search remains unclear and the legal protections it
provides for the enormous amounts of personal data stored
online are far too low.
Finally, Congress should encourage the FTC to investigate
how technology can be harnessed to give consumers better
control over their online information. The lack of effective
controls and the difficulty that consumers have in exercising
choice about their participation in online tracking and
targeting was the motivation behind the ``Do Not Track'' list
idea proposed by CDT and nine other consumer and privacy
groups.\30\ Although the proposal has been controversial, the
idea behind Do Not Track is both simple and important: provide
consumers with an easy-to-use, technology-neutral, persistent
way to opt out of behavioral advertising. Congress should
promote further study of this idea and other innovative ways to
put consumers in control of their information.
---------------------------------------------------------------------------
\30\ See Pam Dixon et al, Consumer Rights and Protections in the
Behavioral Advertising Sector (Oct. 2007), http://www.cdt.org/privacy/
20071031consumerprotectionsbehavioral.pdf.
---------------------------------------------------------------------------
VIII. Conclusion
I would like to thank the Committee again for holding this
important hearing. We believe that Congress has a critical role to play
in ensuring that privacy is protected in an increasingly complex online
advertising environment. CDT looks forward to working with the
Committee as it pursues these issues further.
Appendix A
Simplified Illustration of a Traditional Online Ad Network
Figure 1 below shows a simplified version of a traditional online
ad network. Ad networks contract with advertisers on one side and
publishers on the other. They take the ads they receive from
advertisers and match them to open ad spaces on publisher sites.
Figure 1.
Figure 2 shows how an ad network collects data about a consumer's
web activities. When the consumer first visits a publisher site in the
network (SF-hotel-review.com), the ad network places a cookie with a
unique ID (12345) on the consumer's computer. When the user
subsequently visits other publisher sites in the network (including
dogzblogs.com and social-network.net), the cookie containing the ID is
automatically transmitted to the ad network. This allows the ad network
to keep track of what sites the consumer has visited and build a
behavioral profile based on that information, linked to the cookie ID.
Figure 2.
Appendix B
An Overview of the Federal Wiretap Act, Electronic Communications
Privacy Act, and State Two-Party Consent Laws of Relevance to
the NebuAd System and Other Uses of Internet Traffic Content
from ISPs for Behavioral Advertising--July 8, 2008
Much of the content on the Internet (just like content in
newspapers, broadcast TV, radio and cable) is supported in whole or
part by advertising revenue. The Internet offers special opportunities
to target ads based on the expressed or inferred interests of the
individual user. There are various models for delivering targeted ads
online. These range from the purely contextual (everyone who visits a
travel site sees the same airline ad) to models that involve compiling
information about the online behavior of individual Internet users, to
be used in serving them advertisements. For years, websites have
entered into agreements with advertising networks to use ``cookies'' to
track individual users across websites in order to compile profiles.
This approach has always been, and remains, a source of privacy
concern, in part because the conduct usually occurs unbeknownst to most
Internet users. Recent developments, including the mergers between
online service providers and some of the largest online advertising
networks, have heightened these concerns. The Center for Democracy and
Technology has been conducting a major project on behavioral
advertising, in which we have been researching behavioral advertising
practices, consulting with Internet companies and privacy advocates,
developing policy proposals, filing extensive comments at the FTC, and
analyzing industry self-regulatory guidelines.
This memo focuses on the implications of a specific approach to
behavioral advertising being considered by Internet advertising
networks and Internet Service Providers (ISPs). This new approach
involves copying and inspecting the content of each individual's
Internet activity with the cooperation of his or her ISP.\31\ Under
this new model, an advertising network strikes a deal with an ISP, and
the ISP allows the network to copy the contents of the individual Web
traffic streams of each of the ISP's customers. The advertising network
analyzes the content of these traffic streams in order to create a
record of each individual's online behaviors and interests. Later, as
customers of the ISP surf the Web and visit sites where the advertising
network has purchased advertising space, they see ads targeted based on
their previous Internet behavior.
---------------------------------------------------------------------------
\31\ See, e.g., Peter Whoriskey, Every Click You Make, Wash. Post
(Apr. 3, 2008), http://www.washingtonpost.com/wpdyn/content/article/
2008/04/03/AR2008040304052.html?nav=hc
module; Saul Hansell, I.S.P. Tracking: The Mother of All Privacy
Battles, N.Y. Times: Bits Blog (Mar. 20, 2008), http://
bits.blogs.nytimes.com/2008/03/20/isp-tracking-the-mother-of-all-
privacy-battles/?scp=1-b&sq= the+mother+of+all+privacy+battles&st=nyt.
---------------------------------------------------------------------------
NebuAd is one such advertising network company operating in the
United States. In the past few months, it has come to light that NebuAd
was planning to partner with Charter Communications, a cable broadband
ISP, to conduct trials of the NebuAd behavioral advertising technology.
Several other smaller ISPs, such as Wide Open West (WOW!), CenturyTel,
Embarq, and Knology, have also announced plans with NebuAd to trial or
deploy its behavioral advertising technology. In response to concerns
raised by subscribers, privacy advocates, and policymakers, Charter,
CenturyTel and Embarq have delayed these plans, but NebuAd and other
similar companies are continuing to seek new ISP partners.
The use of Internet traffic content from ISPs for behavioral
advertising is different from the ``cookie''-based model in significant
ways and raises unique concerns.\32\ Among other differences, it copies
all or substantially all Web transactions, including visits to sites
that do not use cookies. Thus, it may capture not only commercial
activity, but also visits to political, advocacy, or religious sites or
other non-commercial sites that do not use cookies.
---------------------------------------------------------------------------
\32\ Privacy concerns also apply to advertising-based models that
have been developed for services, such as e-mail, that ride over ISP
networks. See CDT Policy Post 10.6, Google GMail Highlights General
Privacy Concerns, (Apr. 12, 2004), http://www.cdt.org/publications/
policyposts/2004/6 (recommending express prior opt-in for advertising-
based e-mail service).
---------------------------------------------------------------------------
In this memo, we conclude that the use of Internet traffic content
from ISPs may run afoul of Federal wiretap laws unless the activity is
conducted with the consent of the subscriber.\33\ To be effective, such
consent should not be buried in terms of service and should not be
inferred from a mailed notice. We recommend prior, express consent, but
we do not offer here any detailed recommendations on how to obtain such
consent in an ISP context. Also, we note that that the California law
requiring consent of all the parties to a communication has been
applied by the state Supreme Court to the monitoring of telephone calls
when the monitoring is done at a facility outside California. The
California law so far has not been applied to Internet communications
and it is unclear whether it would apply specifically to the copying of
communications as conducted for behavioral monitoring purposes, but if
it or another state's all-party consent rule were applied to use of
Internet traffic for behavioral profiling, it would seem to pose an
insurmountable barrier to the practice.
---------------------------------------------------------------------------
\33\ Additional questions have been raised under the Cable
Communications Policy Act. See Rep. Edward Markey and Rep. Joe Barton,
Letter to Charter Communications CEO in Regards to the Charter-NebuAd
Data Collection Scheme (May 2008), http://markey.house.gov/docs/
telecomm/ letter_charter_comm_privacy.pdf. In this memo, we focus on
issues arising under the Federal Wiretap Act, as amended by the
Electronic Communications Privacy Act.
---------------------------------------------------------------------------
I. Wiretap Act
A. Service Providers Cannot ``Divulge'' The Contents of Subscriber
Communications, Except Pursuant to Limited Exceptions
The Federal Wiretap Act, as amended by the Electronic
Communications Privacy Act, protects the privacy of wire, oral, and
electronic communications.\34\ ``[E]lectronic communication'' is
defined as ``any transfer of signs, signals, writing, images, sounds,
data, or intelligence of any nature transmitted in whole or in part by
a wire, radio, electromagnetic, photoelectronic or photooptical system
. . ..'' \35\ Web browsing and other Internet communications are
clearly electronic communications protected by the Wiretap Act.
---------------------------------------------------------------------------
\34\ 18 U.S.C. 2510-2522.
\35\ Id. 2510(12).
---------------------------------------------------------------------------
In language pertinent to the model under consideration, 2511(3)
of the Act states that ``a person or entity providing an electronic
communication service to the pubic shall not intentionally divulge the
contents of any communications . . . while in transmission on that
service to any person or entity other than an addressee or intended
recipient. . . .'' \36\
---------------------------------------------------------------------------
\36\ Id. 2511(3)(a). Lest there be any argument that the
disclosure does not occur while the communications are ``in
transmission,'' we note that the Stored Communications Act (SCA) states
that ``a person or entity providing an electronic communication service
to the public shall not knowingly divulge to any person or entity the
contents of a communication while in electronic storage by that
service.'' Id. 2702(a)(1). We do not comment further here on the SCA
because, in our judgment, the approach that has been described so far
clearly involves the divulging of communications ``while in
transmission.''
---------------------------------------------------------------------------
There are exceptions to this prohibition on disclosure, two of
which may be relevant here. One exception specifies that ``[i]t shall
not be unlawful under this chapter for an . . . electronic
communication service, whose facilities are used in the transmission of
a[n] . . . electronic communication, to intercept, disclose, or use
that communication in the normal course of his employment while engaged
in any activity which is a necessary incident to the rendition of his
service or to the protection of the rights or property of the provider
of that service.'' \37\ We will refer to this as the ``necessary
incident'' exception. The second exception is for disclosures with the
consent of one of the parties.\38\ We will discuss both exceptions
below. We conclude that only the consent exception applies to the
disclosure of subscriber content for behavioral advertising, and we
will discuss preliminarily what ``consent'' would mean in this context.
---------------------------------------------------------------------------
\37\ Id. 2511(2)(a)(i) (emphasis added). This analysis focuses on
the capture of electronic communications and definitions are abridged
accordingly.
\38\ Id. 2511(3)(b)(ii).
---------------------------------------------------------------------------
B. With Limited Exceptions, Interception Is Also Prohibited
The Wiretap Act regulates the ``interception'' of electronic
communications. The Act defines ``intercept'' as the ``acquisition of
the contents of any . . . electronic . . . communication through the
use of any electronic, mechanical, or other device.'' \39\
---------------------------------------------------------------------------
\39\ Id. 2510(4).
---------------------------------------------------------------------------
The Wiretap Act broadly bars all intentional interception of
electronic communications.\40\ The Act enumerates specific exceptions
to this prohibition.\41\ Law enforcement officers, for example, are
authorized to conduct interceptions pursuant to a court order. For ISPs
and other service providers, there are three exceptions that might be
relevant. Two we have mentioned already: the ``necessary incident''
exception and a consent exception.\42\
---------------------------------------------------------------------------
\40\ Id. 2511(1).
\41\ Id. 2511(2).
\42\ Separate from the consent provision for disclosure, the
consent exception for interception is set forth in 18 U.S.C.
2511(2)(d): ``It shall not be unlawful under this chapter for a person
not acting under color of law to intercept a[n] . . . electronic
communication where such person is a party to the communication or
where one of the parties to the communication has given prior consent
to such interception. . . .''
---------------------------------------------------------------------------
A third exception, applicable to interception but not to
disclosure, arises from the definition of ``intercept,'' which is
defined as acquisition by an ``electronic, mechanical, or other
device,'' which in turn is defined as ``any device or apparatus which
can be used to intercept a[n] . . . electronic communication other
than--(a) any telephone or telegraph instrument, equipment or facility,
or any component thereof . . . (ii) being used by a provider of . . .
electronic communication service in the ordinary course of its
business. . . .'' \43\ This provision thus serves to limit the
definition of ``intercept,'' providing what is sometimes called the
``telephone extension'' exception, but which we will call the
``business use'' exception.
---------------------------------------------------------------------------
\43\ Id. 2510(5) (emphasis added).
---------------------------------------------------------------------------
C. The Copying of Internet Content for Disclosure to Advertising
Networks Constitutes Interception
When an ISP copies a customer's communications or allows them to be
copied by an advertising network, those communications have undoubtedly
been ``intercept[ed].'' \44\ Therefore, unless an exception applies, it
seems likely that placing a device on an ISP's network and using it to
copy communications for use in developing advertising profiles would
constitute illegal interception under 2511(1)(a); similarly, the
disclosure or use of the intercepted communications would run afoul of
2511(1)(c) or 2511(1)(d), respectively.
---------------------------------------------------------------------------
\44\ See, e.g., United States v. Rodriguez, 968 F.2d 130, 136 (2d
Cir. 1992) (holding in context of telephone communications that ``when
the contents of a wire communication are captured or redirected in any
way, an interception occurs at that time'' and that ``[r]edirection
presupposes interception''); In re State Police Litig., 888 F. Supp.
1235, 1267 (D. Conn. 1995) (stating in context of telephone
communications that ``it is the act of diverting, and not the act of
listening, that constitutes an `interception' '').
---------------------------------------------------------------------------
D. The ``Necessary Incident'' Exception Probably Does Not Permit the
Interception or Disclosure of Communications for Behavioral
Advertising Purposes
The Wiretap Act permits interception of electronic communications
when the activity takes place as ``a necessary incident to the
rendition of [the ISP's] service or to the protection of the rights or
property of the provider of that service.'' \45\ The latter prong
covers anti-spam and anti-virus monitoring and filtering and various
anti-fraud activities, but cannot be extended to advertising
activities, which, while they may enhance the service provider's
revenue, do not ``protect'' its rights. Courts have construed the
``necessary incident'' prong quite strictly, requiring a service
provider to show that it must engage in the activity in order to carry
out its business.\46\ It is unlikely that the copying, diversion, or
disclosure of Internet traffic content for behavioral advertising would
be construed as a ``necessary incident'' to an ISP's business.
Conceivably, an ISP could argue that its business included copying its
subscribers communications and providing them to third parties for
purposes of placing advertisements on websites unaffiliated with the
ISP, but the ISP would probably have to state that that business
existed and get the express agreement of its customers that they were
subscribing to that business as well as the basic business of Internet
access, which leads anyhow to the consent model that we conclude is
necessary.
---------------------------------------------------------------------------
\45\ 18 U.S.C. 2511(2)(a)(i).
\46\ See United States v. Councilman, 418 F.3d 67, 82 (1st Cir.
2005) (en banc) (holding that service provider's capture of e-mails to
gain commercial advantage ``clearly'' was not within service provider
exception); Berry v. Funk, 146 F.3d 1003, 1010 (D.C. Cir. 1998)
(holding in context of telephone communications that switchboard
operators' overhearing of a few moments of phone call to ensure call
went through is a ``necessary incident,'' but anything more is outside
service provider exception).
---------------------------------------------------------------------------
E. While It Is Unclear Whether the ``Business Use'' Exception Would
Apply to the Use of a Device Installed or Controlled by a Party
Other than the Service Provider, the Exception Does Not Apply
to the Prohibition Against Divulging a Subscriber's
Communications
The ``business use'' exception, 2510(5)(a), constricts the
definition of ``device'' and thereby narrows the definition of
``intercept'' in the Wiretap Act. There are two questions involved in
assessing applicability of this exception to the use of Internet
traffic content for behavioral advertising: (1) whether the device that
copies the content for delivery to the advertising network constitutes
a ``telephone or telegraph instrument, equipment or facility, or any
component thereof,'' and (2) whether an ISP's use of the device would
be within the ``ordinary course of its business.''
We will discuss the ``business use'' exception at some length,
because there has been considerable discussion already about whether
copying of an ISP subscriber's communications for behavioral
advertising is an ``interception'' under 2511(1) of the Wiretap Act.
However, even if the business use exception applied, an ISP would only
avoid liability for the interception of electronic communications. It
would still be prohibited from divulging the communications of its
customers to an advertising network under the separate section of the
Wiretap Act, 2511(3), which states that a service provider ``shall
not intentionally divulge the contents of any communication . . . while
in transmission on that service to any person or entity other than an
addressee or intended recipient. . . .'' \47\ The business use
exception does not apply to this prohibition against divulging.\48\
---------------------------------------------------------------------------
\47\ 18 U.S.C. 2511(3)(a).
\48\ By adopting two different exceptions--``necessary incident''
and ``ordinary course''--Congress apparently meant them to have
different meanings. Based on our reading of the cases, the necessary
incident exception is narrower than the ordinary course exception. It
is significant that the ``necessary incident'' exception applies to
both interception and disclosure while the ``ordinary course''
exception is applicable only to interception. This suggests that
Congress meant to allow service providers broader latitude in examining
(that is, ``intercepting'' or ``using'') subscriber communications so
long as they did not disclose the communications to third parties. This
permits providers to conduct a range of in-house maintenance and
service quality functions that do not involve disclosing communications
to third parties.
---------------------------------------------------------------------------
At first glance, it would seem that the business use exception is
inapplicable to the facilities of an ISP because the exception applies
only to a ``telephone or telegraph instrument, equipment or facility,
or any component thereof.'' However, the courts have recognized that
ECPA was motivated in part by the ``dramatic changes in new computer
and telecommunications technologies'' \49\ and therefore was intended
to make the Wiretap Act largely neutral with respect to its treatment
of various communications technologies. The Second Circuit, for
example, concluded in a related context that the term ``telephone''
should broadly include the ``instruments, equipment and facilities that
ISPs use to transmit e-mail.'' \50\ Therefore, as a general matter, it
should be assumed that the business use exception is available to ISPs.
---------------------------------------------------------------------------
\49\ S. Rep. No. 99-541, at 1 (1986), reprinted in 1986
U.S.C.C.A.N. 3555.
\50\ Hall v. Earthlink Network, Inc., 396 F.3d 500, 505 (2d Cir.
2005) (quoting S. Rep. No. 99-541 at 8).
---------------------------------------------------------------------------
However, it is not certain that the device used to copy and divert
content for behavioral advertising would be considered to be a
component of the service provider's equipment or facilities. In some of
the behavioral advertising implementations that have been described,
the monitoring device or process is not developed or controlled by the
ISP but rather by the advertising network.
The second question is whether an ISP's use of a device to copy
traffic content for behavioral advertising falls within the ``ordinary
course of its business.'' There are a number of cases interpreting this
exception, but none of them clearly addresses a situation where a
service provider is copying all of the communications of its customers.
Many of the cases arise in situations where employers are monitoring
the calls of their employees for purposes of supervision and quality
assurance. ``These cases have narrowly construed the phrase `ordinary
course of business.' '' \51\ Often such cases also involve notice to
the employees and implied consent.\52\ One court has stated that, even
if an entity could satisfy the business use exception, notice to one of
the parties being monitored would be required.\53\ Other cases involve
the monitoring of prisoners.
---------------------------------------------------------------------------
\51\ United States v. Murdock, 63 F.3d 1391. 1396 (6th Cir 1995).
\52\ E.g., James v. Newspaper Agency Corp., 591 F.2d 579 (10th Cir.
1979).
\53\ See, e.g., Adams v. City of Battle Creek, 250 F.3d 980, 984
(6th Cir. 2001).
---------------------------------------------------------------------------
Some cases have interpreted ``ordinary course'' to mean anything
that is used in ``normal'' operations. The D.C. Circuit, for instance,
has suggested that monitoring ``undertaken normally'' qualifies as
being within the ``ordinary course of business.'' \54\ In the context
of law enforcement taping of the phone calls of prisoners, the Ninth
and Tenth Circuits have concluded that something is in the ``ordinary
course'' if it is done routinely and consistently.\55\ It might be that
courts would give equal or greater latitude to service providers in
monitoring their networks than they would give to mere subscribers or
users.
---------------------------------------------------------------------------
\54\ Berry v. Funk, 146 F.3d 1003, 1009 (D.C. Cir. 1998) (workplace
monitoring).
\55\ See United States v. Van Poyck, 77 F.3d 285, 292 (9th Cir.
1996); United States v. Gangi, 57 Fed. Appx. 809, 814 (10th Cir. 2003).
---------------------------------------------------------------------------
Other circuit courts have used a more limited interpretation,
concluding that ``ordinary course'' only applies if the device is being
used to intercept communications for ``legitimate business reasons.''
\56\ Although the courts have not been entirely clear as to what that
means, some have suggested that it is much closer to necessity than to
mere profit motive.\57\ One frequently-cited case explicitly holds that
the business use exception does not broadly encompass a company's
financial or other motivations: ``The phrase `in the ordinary course of
business' cannot be expanded to mean anything that interests a
company.'' \58\
---------------------------------------------------------------------------
\56\ See Arias v. Mutual Central Alarm Serv., Inc., 202 F.3d 553,
560 (2d Cir. 2000) (monitoring calls to an central alarm monitoring
service).
\57\ See id. (concluding that alarm company had legitimate reasons
to tap all calls because such businesses ``are the repositories of
extremely sensitive security information, including information that
could facilitate access to their customers' premises''); see also First
v. Stark County Board of Comm'rs, 234 F.3d 1268, at *4 (6th Cir. 2000)
(table disposition).
\58\ Watkins v. L.M. Berry & Co., 704 F.2d 577, 582 (11th Cir.
1983). Watkins states: ``We hold that a personal call may not be
intercepted in the ordinary course of business under the exemption in
section 2510(5)(a)(i), except to the extent necessary to guard against
unauthorized use of the telephone or to determine whether a call is
personal or not. In other words, a personal call may be intercepted in
the ordinary course of business to determine its nature but never its
contents.'' 704 F.2d at 583. This language supports the conclusion that
the business use exception could not cover wholesale interception of
ISP traffic, no more than switchboard operators can perform wholesale
monitoring of telephone traffic.
---------------------------------------------------------------------------
Normal principles of statutory interpretation would require that
some independent weight be given to the word ``ordinary,'' so that the
exception does not encompass anything done for business purposes. It is
unclear, however, how much weight courts would give to the word
``ordinary'' in a rapidly changing market. It does not seem that the
phrase ``ordinary course of business'' should preclude innovation, but
courts might refer to past practices and normal expectations
surrounding a line of business and specifically might look to what
customers have come to expect.
Viewed one way, it is hard to see how the copying of content for
behavioral advertising is part of the ``ordinary course of business''
of an ISP. After all, the ISP is not the one that will be using the
content to develop profiles of its customers; the profiling is done by
the advertising network, which does not even disclose to the ISP the
profiles of its own subscribers. (The profiles are proprietary to the
advertising network and it is careful not to disclose them to anyone.)
Very few (if any) of the ads that are placed using the profiles will be
ads for the ISP's services; they will be ads for products and services
completely unrelated to the ISP's ``ordinary course of business.''
Moreover, the ads will be placed on websites having no affiliation with
the ISP. On the other hand, the ISP could argue that part of its
business model--part of what keeps its rates low--is deriving revenue
from its partnership with advertising networks.
The legislative histories of the Wiretap Act and ECPA weigh against
a broad reading of the business use exception. Through these laws,
Congress intended to create a statutory regime generally affording
strong protection to electronic communications. Congress included
limited, specific and detailed exceptions for law enforcement access to
communications, and other limited, specific and detailed exceptions to
allow companies providing electronic communications service to conduct
ordinary system maintenance and operational activities. Congress gave
especially high protection to communications content. If the business
use exception can apply any time an ISP identifies a new revenue stream
that can be tapped though use of its customers' communications, this
careful statutory scheme would be seriously undermined.
F. The Consent Exception: The Context Weighs Heavily in Favor of
Affirmative, Opt-In Consent from ISP Subscribers
Consent is an explicit exception both to the prohibition against
intercepting electronic communications under the Wiretap Act and to the
Act's prohibition against disclosing subscriber communications. The key
question is: How should consent be obtained for use of Internet traffic
content for behavioral advertising? Courts have held in telephone
monitoring cases under the Wiretap Act that consent can be implied, but
there are relatively few cases specifically addressing consent and
electronic communications. However, in cases involving telephone
monitoring, one circuit court has stated that consent under the Wiretap
Act ``is not to be cavalierly implied.'' \59\ Another circuit court has
noted that consent ``should not casually be inferred'' \60\ and that
consent must be ``actual,'' not ``constructive.'' \61\ Yet another
circuit court has stated: ``Without actual notice, consent can only be
implied when the surrounding circumstances convincingly show that the
party knew about and consented to the interception.'' \62\ Furthermore,
``knowledge of the capability of monitoring alone cannot be considered
implied consent.'' \63\ The cases where consent has been implied
involve very explicit notice; many of them involve the monitoring of
prisoners' phone calls.\64\
---------------------------------------------------------------------------
\59\ Watkins. 704 F.2d at 581 (``Consent under title III is not to
be cavalierly implied. Title III expresses a strong purpose to protect
individual privacy by strictly limiting the occasions on which
interception may lawfully take place.'').
\60\ Griggs-Ryan v. Smith, 904 F.2d 112, 117 (1st Cir. 1990).
\61\ In re Pharmatrak, Inc. Privacy Litig., 329 F.3d 9, 20 (1st
Cir. 2003); see also United States v. Corona-Chavez, 328 F.3d 974, 978
(8th Cir. 2003).
\62\ Berry v. Funk, 146 F.3d 1003, 1011 (D.C. Cir. 1998) (internal
quotation omitted).
\63\ Watkins, 704 F.2d at 581; see also Deal v. Spears, 980 F.2d
1153, 1157 (8th Cir. 1992) (holding that consent not implied when
individual is aware only that monitoring might occur, rather than
knowing monitoring is occurring).
\64\ ``The circumstances relevant to an implication of consent will
vary from case to case, but the compendium will ordinarily include
language or acts which tend to prove (or disprove) that a party knows
of, or assents to, encroachments on the routine expectation that
conversations are private. And the ultimate determination must proceed
in light of the prophylactic purpose of Title III--a purpose which
suggests that consent should not casually be inferred.'' Griggs-Ryan,
904 F.2d at 117.
---------------------------------------------------------------------------
Consent is context-based. It is one thing to imply consent in the
context of a prison or a workplace, where notice may be presented as
part of the daily log-in process. It is quite another to imply it in
the context of ordinary Internet usage by residential subscribers, who,
by definition, are using the service for personal and often highly
sensitive communications. Continued use of a service after a mailed
notice might not be enough to constitute consent. Certainly, mailing
notification to the bill payer is probably insufficient to put all
members of the household who share the Internet connection on notice.
Thus, it seems that an assertion of implied consent, whether or not
users are provided an opportunity to opt out of the system, would most
likely not satisfy the consent exception for the type of interception
or disclosure under consideration here. Express prior consent (opt-in
consent) is clearly preferable and may be required. While meaningful
opt-in consent would be sufficient, courts would likely be skeptical of
an opt-in consisting merely of a click-through agreement--i.e., a set
of terms that a user agrees to by clicking an on-screen button--if it
displays characteristics typical of such agreements, such as a large
amount of text displayed in a small box, no requirement that the user
scroll through the entire agreement, or the opt-in provision buried
among other terms of service.\65\
---------------------------------------------------------------------------
\65\ See, e.g., Specht v. Netscape Commc'ns Corp., 306 F.3d 17 (2d
Cir. 2002) (rejecting online arbitration agreement because, among other
things, site permitted customer to download product without having
scrolled down to arbitration clause and agreement button said only
``Download''); United States v. Lanoue, 71 F.3d 966, 981 (1st Cir.
1995) (``Deficient notice will almost always defeat a claim of implied
consent.'').
---------------------------------------------------------------------------
In regards to consent, the model under discussion here is
distinguishable from the use of ``cookies,'' which were found to be
permissible by a Federal district court in a 2001 case involving
DoubleClick.\66\ In that case, the websites participating in the
DoubleClick advertising network were found to be parties to the
communications of the Internet users who visited those sites. As
parties to the communications, the websites could consent to the use of
the cookies to collect information about those communications. Here, of
course, the ISPs are not parties to the communications being monitored
and the interception or disclosure encompasses communications with
sites that are not members of the advertising network. Therefore, the
source of consent must be the IPS's individual subscribers, as it would
be impossible to obtain consent from every single website that every
subscriber may conceivably visit.
---------------------------------------------------------------------------
\66\ In re DoubleClick Inc. Privacy Litig., 154 F.Supp.2d 497
(S.D.N.Y. 2001).
---------------------------------------------------------------------------
II. State Laws Requiring Two-Party Consent to Interception
A. Summary
In addition to the Federal Wiretap Act, a majority of states have
their own wiretap laws, which can be more stringent than the Federal
law. Most significantly, twelve states \67\ require all parties to
consent to the interception or recording of certain types of
communications when such interception is done by a private party not
under the color of law.
---------------------------------------------------------------------------
\67\ The twelve states are California, Connecticut, Florida,
Illinois, Maryland, Massachusetts, Michigan, Montana, Nevada, New
Hampshire, Pennsylvania, and Washington.
---------------------------------------------------------------------------
In several of these states--for example, Connecticut--the all-party
consent requirement applies only to the recording of oral
conversations. In others, the all-party consent rule extends to both
voice and data communications. For example, Florida's Security of
Communications Act makes it a felony for any individual to intercept,
disclose, or use any wire, oral, or electronic communication, unless
that person has obtained the prior consent of all parties \68\
Similarly, the Illinois statute on criminal eavesdropping prohibits a
person from ``intercept[ing], retain[ing], or transcrib[ing an]
electronic communication unless he does so . . . with the consent of
all of the parties to such . . . electronic communication.'' \69\
---------------------------------------------------------------------------
\68\ Fla. Stat. 934.03(1).
\69\ Ill. Comp Stat. 5/14-1(a)(1).
---------------------------------------------------------------------------
The most important all-party consent law may be California's,
because the California Supreme Court held in 2006 that the law can be
applied to activity occurring outside the state.
B. California
The 1967 California Invasion of Privacy Act makes criminally liable
any individual who ``intentionally taps, or makes any unauthorized
connection . . . or who willfully and without the consent of all
parties to the communication . . . reads, or attempts to read, or to
learn the contents or meaning of any message . . . or communication
while the same is in transit or passing over any wire, line, or cable,
or is being sent from, or received at any place'' in California.\70\ It
also establishes liability for any individual ``who uses, or attempts
to use, in any manner . . . any information so obtained'' or who aids
any person in doing the same.\71\ The law has a separate section
creating liability for any person eavesdropping upon or recording a
confidential communication ``intentionally and without the consent of
all parties,'' whether the parties are present in the same location or
communicating over telegraph, telephone, or other device (except a
radio).\72\
---------------------------------------------------------------------------
\70\ Cal. Pen. Code 631(a).
\71\ Id.
\72\ Id. 632(a). The statute explicitly excludes radio
communications from the category of confidential communications.
---------------------------------------------------------------------------
Consent can be implied only in very limited circumstances. The
California State Court of Appeals held in People v. Garber that a
subscriber to a telephone system is deemed to have consented to the
telephone company's monitoring of his calls if he uses the system in a
manner that reasonably justifies the company's belief that he is
violating his subscription rights, and even then the company may only
monitor his calls to the extent necessary for the investigation.\73\ An
individual can maintain an objectively reasonable expectation of
privacy by explicitly withholding consent for a tape recording, even if
the other party has indicated an intention to record the
communication.\74\
---------------------------------------------------------------------------
\73\ 275 Cal. App. 2d 119 (Cal. App. 1st Dist. 1969).
\74\ Nissan Motor Co. v. Nissan Computer Corp., 180 F. Supp. 2d
1089 (C.D. Cal. 2002).
---------------------------------------------------------------------------
In Kearney v. Salomon Smith Barney, Inc., the state Supreme Court
addressed the conflict between the California all-party consent
standard and Georgia's wiretap law, which is modeled after the Federal
one-party standard.\75\ It held that, where a Georgia firm recorded
calls made from its Georgia office to residents in California, the
California law applied. The court said that it would be unfair to
impose damages on the Georgia firm, but prospectively the case
effectively required out-of-state firms having telephone communications
with people in California to announce to all parties at the outset
their intent to record a communication. Clear notice and implied
consent are sufficient. ``If, after being so advised, another party
does not wish to participate in the conversation, he or she simply may
decline to continue the communication.'' \76\
---------------------------------------------------------------------------
\75\ 39 Cal. 4th 95 (2006).
\76\ Id. At 118.
---------------------------------------------------------------------------
C. The Implications of Kearney
The Kearney case arose in the context of telephone monitoring, and
there is a remarkable lack of case law addressing whether the
California statute applies to Internet communications. If it does, or
if there is one other state that applies its all-party consent rule to
conduct affecting Internet communications across state lines, then no
practical form of opt-in, no matter how robust, would save the practice
of copying Internet content for behavioral advertising. That is, even
if the ISP only copies the communications of those subscribers that
consent, and the monitoring occurs only inside a one-party consent
state, as soon as one of those customers has a communication with a
non-consenting person (or website) in an all-party consent state that
applies its rule to interceptions occurring outside the state, the ISP
would seem to be in jeopardy. The ISP could not conceivably obtain
consent from every person and website in the all-party consent state.
Nor could it identify (for the purpose of obtaining consent) which
people or websites its opted-in subscribers would want to communicate
with in advance of those communications occurring.
A countervailing argument could be made that an all-party consent
rule is not applicable to the behavioral advertising model, since the
process only copies or divulges one half of the communication, namely
the half from the consenting subscriber.
III. Conclusion
The practice that has been described to us, whereby an ISP may
enter into an agreement with an advertising network to copy and analyze
the traffic content of the ISP's customers, poses serious questions
under the Federal Wiretap Act. It seems that the disclosure of a
subscriber's communications is prohibited without consent. In addition,
especially where the copying is achieved by a device owned or
controlled by the advertising network, the copying of the contents of
subscriber communications seems to be, in the absence of consent, a
prohibited interception. Affirmative express consent, and a cessation
of copying upon withdrawal of consent, would probably save such
practices under Federal law, but there may be state laws requiring all-
party consent that would be more difficult to satisfy.
Senator Dorgan. Ms. Harris, thank you very much. We
appreciate your testimony.
Mr. Chris Kelly is the Chief Privacy Officer for Facebook
Incorporated. Mr. Kelly, you may proceed.
STATEMENT OF CHRIS KELLY, CHIEF PRIVACY OFFICER, FACEBOOK, INC.
Mr. Kelly. Thank you very much, Chairman Dorgan and Members
of the Committee, for the opportunity to address the Committee
about the important privacy matters facing the online
advertising industry.
I am Chris Kelly, the Chief Privacy Officer of Facebook, a
social service on the Internet that serves more than 80 million
active users, about 30 million of whom are in the United
States.
Facebook aims to create social value by empowering people
to share their lives and experiences with the people they care
about. From the founding of the company in a dorm room in 2004
until today, Facebook's privacy settings have given users
control over who has access to their personal information by
allowing them to choose the friends they accept and the
networks they join.
We are dedicated to developing advertising that is relevant
and personal and to transparency with our users about how we
use their information in the advertising context. We are
pleased to discuss both Facebook's general approach to privacy
and how these principles have been implemented in advertising
provided by Facebook.
With many mainstream media reports focusing on privacy
concerns about social networking sites, we first want to
clarify how our site differs from most. Though we will not
always address user concerns perfectly--no site can--Facebook
is committed to empowering users to make their own choices
about what information they share and with whom they share it.
The statement that opens our privacy policy, a short, plain
English introduction, is the best place to start this
discussion. It reads: ``We built Facebook to make it easy to
share information with your friends and people around you. We
understand you may not want everyone in the world to have the
information you share on Facebook; that is why we give you
control of your information. Our default privacy settings limit
the information displayed in your profile to your networks and
other reasonable community limitations we tell you about.''
Facebook follows two core principles:
First, you should have control over your personal
information. Facebook helps you share information with your
friends and people around you. You choose what information you
put in your profile, including contact and personal
information, pictures, interests and groups that you join. And
you control the users with whom you share that information
through the privacy settings on the Privacy page.
Two, you should have access to the information that others
want to share. There is an increasing amount of information
available out there, and you may want to know what relates to
you, your friends, and people around you. We want to help you
easily get that information.
Sharing information should be easy. And we want to provide
you with the privacy tools necessary to control how and with
whom you share that information. If you have any questions or
ideas, please send them to [email protected], the e-mail
address that we regularly monitor.
We implement these principles through our friend and
network architectures and through controls that are built into
every one of our innovative products. Contrary to common public
reports, full profile data on Facebook is not even available to
most users on Facebook, let alone all users of the Internet.
Users have extensive and precise controls available to choose
who sees what among their networks and friends, as well as
tools that give them the choice to make a limited set of
information available to search engines and other outside
entities.
The privacy link that appears in the upper right-hand
corner of every Facebook page allows users to make these
choices whenever they are using the site, and everyday use of
the site educates users as to the meaning of privacy controls.
For instance, a user will see regularly that they have access
to the profiles of their friends and those who share a network
with them, but not to profiles of those who are neither friends
nor network members.
In February 2008, Facebook simplified and streamlined its
presentation of privacy settings to users, adopting a common
lock icon throughout the site to denote the presence of a user-
configurable privacy setting. We also introduced the concept of
``Friends Lists'' which, when paired with privacy settings,
allow users to easily configure subsets of their confirmed
friends who may see certain content. We are constantly looking
for means to give users more effective control over their
information and to improve communications with users and the
general public about our privacy architecture so that they can
make their own choices about what they want to reveal.
I want to say a few words about privacy and advertising on
Facebook. It is important to stress in the first instance that
targeting of advertising generally benefits users. But we have
revealed in our privacy policy for nearly 3 years the
following. We have had the following statement present.
``Facebook may use information in your profile without
identifying you as an individual to third parties. We do this
for purposes such as aggregating how many people in a network
like a band or a movie and personalizing advertisements and
promotions so that we can provide you Facebook. We believe this
benefits you. You can know more about the world around you and,
where there are advertisements, they are more likely to be
interesting to you. For example, if you put a favorite movie in
your profile, we might serve you an advertisement highlighting
a screening of a similar one in your town. But we don't tell
the movie company who you are.''
This critical distinction that we embrace in our policies
and practices and that we want users to understand is between
the use of personal information for advertisements in
personally identifiable form and the use, dissemination, or
sharing of information with advertisers in non-personally
identifiable form. Ad targeting that shares or sells personally
identifiable information to advertisers without user control is
fundamentally different from targeting that only gives
advertisers the ability to present their ads based on aggregate
data.
And with that, I see that my time is up. So I look forward
to answering your questions. Thank you.
[The prepared statement of Mr. Kelly follows:]
Prepared Statement of Chris Kelly, Chief Privacy Officer, Facebook,
Inc.
Thank you, Mr. Chairman, for the opportunity to address the
Committee about the important privacy matters facing the online
advertising industry.
I am Chris Kelly, the Chief Privacy Officer of Facebook, a social
service on the Internet that serves more than 80 million active users,
roughly 30 million of whom are in the United States.
Facebook aims to create social value by empowering people to share
their lives and experiences with the people they care about. From the
founding of the company in a dorm room in 2004 to today, Facebook's
privacy settings have given users control over who has access to their
personal information by allowing them to choose the friends they accept
and networks they join.
We are dedicated to developing advertising that is relevant and
personal, and to transparency with our users about how we use their
information in the advertising context. We are pleased to discuss both
Facebook's general approach to privacy and how these principles have
been implemented in advertising provided by Facebook.
With many mainstream media reports focusing on privacy concerns
about ``social networking sites,'' we first want to clarify how our
site differs from most. Though we will not always address user concerns
perfectly--no site can--Facebook is committed to empowering users to
make their own choices about what information they share, and with whom
they share it.
I. Facebook and Privacy
The statement that opens our privacy policy, a short plain-English
introduction, is the best place to start this discussion. It reads:
We built Facebook to make it easy to share information with
your friends and people around you. We understand you may not
want everyone in the world to have the information you share on
Facebook; that is why we give you control of your information.
Our default privacy settings limit the information displayed in
your profile to your networks and other reasonable community
limitations that we tell you about.
Facebook follows two core principles:
1. You should have control over your personal information.
Facebook helps you share information with your friends and
people around you. You choose what information you put in your
profile, including contact and personal information, pictures,
interests and groups you join. And you control the users with
whom you share that information through the privacy settings on
the Privacy page.
2. You should have access to the information others want to
share.
There is an increasing amount of information available out
there, and you may want to know what relates to you, your
friends, and people around you. We want to help you easily get
that information.
Sharing information should be easy. And we want to provide you
with the privacy tools necessary to control how and with whom
you share that information. If you have questions or ideas,
please send them to [email protected].
We implement these principles through our friend and network
architectures, and through controls that are built into every one of
our innovative products. Contrary to common public reports, full
profile data on Facebook isn't even available to most users on
Facebook, let alone all users of the Internet. Users have extensive and
precise controls available to choose who sees what among their networks
and friends, as well as tools that give them the choice to make a
limited set of information available to search engines and other
outside entities.
The ``privacy'' link that appears in the upper-right hand corner of
every Facebook page allows users to make these choices whenever they
are using the site, and everyday use of the site educates users as to
the meanings of privacy controls. For instance, a user will see
regularly that they have access to the profiles of their friends and
those who share a network, but not to the profiles of those who are
neither friends nor network members.
In February 2008, Facebook simplified and streamlined its
presentation of privacy settings to users, adopting a common lock icon
throughout the site to denote the presence of a user-configurable
privacy setting. We also introduced the concept of ``Friends Lists,''
which, when paired with privacy settings, allow users to easily
configure a subset of their confirmed friends who may see certain
content. We are constantly looking for means to give users more
effective control over their information and to improve communications
with users and the general public about our privacy architecture so
they can make their own choices about what they want to reveal.
For instance, we participated in the Federal Trade Commission's
workshop on new advertising technologies, and have been working with
government officials and nongovernmental organizations throughout the
globe. Facebook has also worked productively with state and Federal
officials, as well as law enforcement, to explain our longstanding
strategy to make the Internet safer by promoting responsibility and
identity online, and is currently participating in the state Attorneys
General Internet Safety Technical Task Force.
II. Privacy and Advertising on Facebook
A. Personally Identifiable and Non-Personally Identifiable Information
It is important to stress here in the first instance that targeting
of advertising generally benefits users. Receiving information that is
likely to be relevant, whether paid for by an advertiser or not, leads
to a better online experience. Facebook aims to be transparent with our
users about the fact that advertising is an important source of our
revenue and to explain to them fully the uses of their personal data
they are authorizing by using Facebook. For instance, the following
explanation of how we use information for advertising has been a
prominent part of our privacy policy for nearly 3 years:
Facebook may use information in your profile without
identifying you as an individual to third parties. We do this
for purposes such as aggregating how many people in a network
like a band or movie and personalizing advertisements and
promotions so that we can provide you Facebook. We believe this
benefits you. You can know more about the world around you and,
where there are advertisements, they're more likely to be
interesting to you. For example, if you put a favorite movie in
your profile, we might serve you an advertisement highlighting
a screening of a similar one in your town. But we don't tell
the movie company who you are.
The critical distinction that we embrace in our policies and
practices, and that we want users to understand, is between the use of
personal information for advertisements in personally-identifiable
form, and the use, dissemination, or sharing of information with
advertisers in non-personally-identifiable form. Ad targeting that
shares or sells personal information to advertisers (name, e-mail,
other contact oriented information) without user control is
fundamentally different from targeting that only gives advertisers the
ability to present their ads based on aggregate data. Most Facebook
data is collected transparently in personally identifiable form--users
know they are providing the data about themselves and are not forced to
provide particular information.\1\ Sharing information on the site is
limited by user-established friend relationships and user-selected
networks that determine who has access to that personal information.
Users can see how their data is used given the reactions of their
friends when they update their profiles, upload new photos or videos,
or update their current status.
---------------------------------------------------------------------------
\1\ Currently, only four pieces of data are required to establish
and maintain a Facebook account--e-mail address to provide a unique
login identifier, birthdate to calculate age, name to provide a
standard identifier (our Terms of Use require real name), and gender to
promote the accuracy of grammar through the site infrastructure.
---------------------------------------------------------------------------
On Facebook, then, a feedback loop is established where people know
what they are uploading and receive timely reactions from their
friends, reinforcing the fact they have uploaded identifiable
information. The privacy policy and the users' experiences inform them
of how advertising on the service works--advertising that enables us to
provide the service for free to users is targeted to the expressed
attributes of a profile and presented in the space on the page
allocated for advertising, without granting an advertiser access to any
individual user's profile.
Furthermore, advertising on Facebook is subject to guidelines
designed to avoid deceptive practices, and with special restrictions
and review with respect to any advertising targeted at minors.
I cannot stress strongly enough that Facebook does not authorize
access by the Internet population at large, including advertisers, to
the personally identifiable information that a user willingly uploads
to Facebook. Facebook profiles have extensive user-configurable rules
limiting access to information contained in them. Unless a user decides
otherwise by willingly sharing information with an advertiser--for
instance, through a contest--advertisers may only target advertisements
against non-personally identifiable attributes about a user of Facebook
derived from profile data.
We recognize that other Internet services may take a different
approach to advertisers and the information available to them.
Advertising products that sell personally identifiable information to
advertisers without user permission, that rely on transforming non-
personally identifiable information into personally identifiable
information without robust notice and choice to users, or that rely on
data collection that a user has scant notice of and no control over,
raise fundamentally different privacy concerns. Facebook does not offer
such products today and has no intention of doing so. Advertising
products founded on the principles of transparency and user control,
where data is collected directly from users in personally identifiable
space and targeting is done based on aggregate or characteristic data
in non-personally identifiable space, respect the principle that sits
at the heart of privacy concerns.
B. History of Facebook Ads and Beacon
Perhaps because our site has developed so quickly, we have
sometimes been inartful in communicating with our users and the general
public about our advertising products. It therefore may be fruitful to
provide a brief history of the current Facebook advertising offerings,
including Facebook Ads and Social Ads, as well as the Beacon product
that garnered significant public attention late last year.
In November 2007, Facebook introduced Facebook Ads, which consisted
of both a basic self-service targeting infrastructure based on the non-
personally identifiable use of keywords derived from profile data, and
Social Ads, which allow for the paid promotion of certain interactions
users take online to those users' friends in conjunction with an
advertiser message. The basic targeting infrastructure of Facebook Ads
is quite similar to many other Internet advertising systems, where
media buyers and agencies can purchase guarantees that their
advertisements will run to people who have certain characteristics,
often expressed (as they are in Facebook Ads) in ``keywords,'' or in
demographic categories such as men between 29 and 34.
Social Ads are an innovation in that they allow advertisers to pay
for promotion of certain interactions users take online to those users'
friends. For example, if I become a supporter of a particular political
figure on Facebook, their campaign could pay to promote that fact to
more of my friends than would have been informed of it otherwise
through the Facebook News Feed, and potentially pair a message from the
campaign with it. It is notable first that only my action can trigger a
Social Ad and that Social Ads are only presented to confirmed friends
as opposed to the world at large; there will be no Social Ad generated
noting my action to anyone but a confirmed friend. It is also notable
that in this paid promotion context through Social Ads, an advertiser
is not purchasing and does not have access to users' personal data--
they are only told that a certain number of users have taken relevant
actions and the number of ads generated by those actions.
We introduced at the same time as Facebook Ads a product called
Beacon to allow users to bring actions they take on third-party sites
into Facebook. Our introduction of this product with advertising
technology led many to believe that Beacon was an ad product when it
really was not. Participating third party sites do not pay Facebook to
offer Beacon, nor must a third party site that wants to use Beacon
purchase Facebook Ads. No Facebook user data is sold to or shared with
these third party sites. In most cases, Beacon pertains to non-
commercial actions like the playing of a game or the adding of a recipe
to an online recipe box. In other cases, we and the participating third
party sites experimented with capturing purchases for sharing within a
user's Facebook friend network, obviously a more commercial enterprise.
In both the non-commercial and commercial contexts, we discovered in
the weeks after launch that users felt they did not have adequate
control over the information and how it was being shared with their
friends.
We quickly reached the conclusion that Beacon had inadequate built-
in controls driving user complaints, helped along by an organized
campaign by MoveOn.org to get us to alter the product. We made
significant changes within weeks after its launch to make it a fully
opt-in system. We remain convinced that the goal of helping users share
information about their activities on the web with their friends is
desirable and appreciated. Indeed, a number of services now exist which
attempt to help users in this way. While Beacon was cast in the
mainstream press as an advertising product, it operates fundamentally
as a means to connect, with a user's permission and control, actions
elsewhere on the web with a user's Facebook friend network.
We are currently working on the next generation of Facebook's
interactions with third party websites, called Facebook Connect, to
empower users further to share content and actions with their friends
using the Facebook infrastructure, and are focused on assuring that
proper controls are built into this system.
III. FTC Principles on Behavioral Targeting
Finally, we would like to reinforce our earlier positive public
comments about the Federal Trade Commission's leadership in addressing
privacy concerns about how data is collected and shared online.
As explained above, Facebook Ads are materially different from
behavioral targeting as it is usually discussed, but given our goals of
transparency and user control, the important corollary of ensuring
appropriate security and the goal of providing users notice and choice
with respect to service changes, we applaud the FTC's desire to
establish principles in the online advertising area. We believe the FTC
should expand and enhance the discussion in the principles about the
distinction between personally and non-personally identifiable
information to clarify the need for different treatment of advertising
based on those different types of information. We will continue our
participation in discussion of the principles as they evolve.
Thank you again, Mr. Chairman, for the opportunity to share our
views, and I am happy to answer any questions you may have.
Attachment
Microsoft's Leadership on Consumer Privacy--July 2008
Microsoft has a long-standing commitment to consumer privacy and we
have put that commitment into action. Here are some examples:
Broad Self-regulatory Approach for Online Advertising. Microsoft
recently filed comments with the Federal Trade Commission explaining
the need for a broad self-regulatory privacy approach to online
advertising, noting that all online advertising activities involve data
collection from users and therefore have privacy implications.
Meaningful Online Advertising Principles. In July 2007, Microsoft
announced five fundamental privacy principles for online search and ad
targeting. These principles include commitments to user notice, user
control, search data anonymization, security, and best practices.
Clear and Upfront User Notice. Microsoft was one of the first
companies to develop so-called ``layered'' privacy notices that give
clear and concise bullet-point summaries of our practices and direct
users to a place where they can find more information. We post a link
to this user-friendly privacy notice on every one of our web pages.
Robust User Control. Microsoft has recently deployed a robust
method to enable users to opt out of behavioral advertising.
Specifically, users can now tie their opt-out choice to their Windows
Live ID so their choice can work across multiple computers and be more
persistent (for example, deleting cookies will not erase their opt-out
selection). We also highlight the availability of this opt-out choice
on the first layer of our privacy notice.
Unique Steps To De-Identify Data. Microsoft is unique in our use of
a technical method (known as a one-way cryptographic hash) to separate
search terms from account holders' personal information, such as name,
e-mail address, and phone number, and to keep them separated in a way
that prevents them from being easily recombined. We have also relied on
this method to ensure that we use only data that does not personally
identify individual consumers to serve ads online.
Strict Search Data Anonymization. Microsoft will anonymize all
search data after 18 months, which we believe is an appropriate time-
frame in our circumstances to enable us to maintain and improve the
security, integrity and quality of our services. In addition, unlike
other companies, we will irreversibly remove the entire IP address and
other cross-session identifiers, such as cookies and other machine
identifiers, from search terms after 18 months.
Support for Federal and State Privacy Legislation. Microsoft has
actively supported state legislation that would impose baseline notice,
choice, and security requirements on entities that collect data to
serve online ads. We also were one of the first companies to advocate
for comprehensive Federal privacy legislation in the United States.
Dedicated Privacy Personnel and Processes. Microsoft was one of the
first companies to appoint a chief privacy officer, an action we took
nearly a decade ago, and we currently employ over 40 employees who
focus on privacy full-time, and another 400 who focus on it as part of
their jobs. We have made significant investments in privacy in terms of
dedicated personnel and training and by building robust privacy
standards into our product development and other business processes.
Guidelines for Third Parties. Microsoft is committed to helping
others in industry protect consumers' privacy interests. For example,
we have released a set of privacy guidelines designed to help
developers build meaningful privacy protections into their software
programs and online services.
Consumer Education and Private-Public Sector Partnerships.
Microsoft has taken steps to educate consumers about ways to protect
themselves while online, and we have worked closely with industry
members and law enforcement around the world to identify security
threats, share best practices, and improve our coordinated response to
privacy, security and other Internet safety issues.
Senator Dorgan. Mr. Kelly, thank you very much. We
appreciate your testimony.
We will now hear from Mr. Clyde Wayne Crews, Jr., Vice
President for Policy, Director of Technology Studies at the
Competitive Enterprise Institute. Mr. Crews, welcome.
STATEMENT OF CLYDE WAYNE CREWS, JR., VICE PRESIDENT
FOR POLICY/DIRECTOR OF TECHNOLOGY STUDIES,
COMPETITIVE ENTERPRISE INSTITUTE
Mr. Crews. Thank you very much. Good morning, Mr. Chairman.
I appreciate the opportunity to appear.
Online behavioral marketing is not the devil, but with
emergent technologies like biometrics on the horizon, data
privacy debates like the ones we are facing today are only
going to intensify. Targeted advertising helps fuel today's
flood of information, frictionless e-commerce, and the global
blogger soapbox. It has become cliche to note that the Internet
is one of the most important wealth-creating and democratizing
technologies ever known.
But behavioral marketing stokes privacy fears. Is my data
personally identifiable? Can it become so? Will my identity be
stolen, and if a breach occurs, who is punished?
Note that we were all angry when ads were untargeted spam.
Now that ads are relevant, well, we are still not satisfied.
Behavioral advertising employs heretofore unexploited
capabilities of the Internet, reinforcing the reality that
there is much more to the Internet than the web at any one
juncture. It is only 2008.
User preferences preclude one-size-fits-all privacy policy.
Online some hide behind digital gated communities. Others
parade in front of personal webcams. Privacy is not a thing to
legislate. It is a relationship expressed in countless ways.
Legislation would be complex. If online privacy is regulated,
what about offline? Should the standard be opt-in or opt-out?
Who defines behavioral or sensitive? Should state laws be
preempted? What about noncommercial information collection?
Industry already follows principles like the FTC's proposed
opt-out for sensitive information even when the information is
not personally identifiable, but the rise of the information
society amid a homeland security culture is an unfortunate
coincidence. Blurring of public and private databases
complicates things. Programs like Total Information Awareness,
CAPPSII, and a national ID undermine privacy when data cannot
be confined to an agreed-upon business purpose.
Government often does not need to protect privacy but to
allow it in the first place. One is reminded of the old Peanuts
cartoon of Snoopy sitting on top of the doghouse typing ``Dear
IRS, please remove my name from your mailing list.''
Another old joke goes that if McDonald's were giving away
free Big Macs in exchange for a DNA sample, there would be
lines around the block.
But consumers do care. The Net itself enables collective
consumer discontent, such as blog backlashes we have seen
against companies. The result: firms alter their information
handling procedures without law. Consumers can also avoid
certain sites or use Anonymizer or Scroogle or TrackMeNot or a
virtual private network. Choice mandates are not persuasive
when choice is increasingly the default.
This debate's fury implies that real market opportunities
exist in providing online anonymity. A marketer does not
necessarily want to know who you are but how somebody like you
acts. No one in a free market is really lucky enough to self-
regulate as FTC puts it. Firms are regulated by consumer fury,
rivals, by Wall Street, by intolerant investors. There is no
such thing as no regulation. The choice we face is between
political discipline or competitive discipline in an impatient
market. Even companies on the frontier of behavioral
advertising like Phorm and NebuAd face discipline. One's
sympathies there are going to depend upon the ownership status
one accords to a web page. But today's web page is not what
tomorrow's web page is going to be with information and ads
coming into the page from numerous sources.
Privacy standards best thrive as a war between computer
scientists. Marketing to an unidentified customer is today's
happy goal, but at the very same time, there is great value in
technologies that prevent others from posing as us. That is one
reason the use of personally identifiable data should not be
ruled out altogether.
Meanwhile, we need improved cyberinsurance products and
enhanced liability products to evolve online. Regulating can
short-circuit such market innovations. The private sector needs
practice for the really difficult cases like the emergence of
biometrics.
Privacy policies are already legally binding. Thus, a more
fitting Federal agenda would target identity theft and computer
crime and enforce privacy policies and stay neutral on computer
science, keep compulsory databases separate from private ones,
stabilize Government's own insecure networks, and avoid
interventions like data retention that undermine security
guarantees.
To protect consumers online, we must consciously avoid
entrenching regulations such that effective private
alternatives and institutions, however warranted, simply cannot
emerge. Online marketers are today's battered business bureau,
but they need battering by competitive discipline, not just
legislation.
Thank you very much.
[The prepared statement of Mr. Crews follows:]
Prepared Statement of Clyde Wayne Crews, Jr., Vice President for
Policy/Director of Technology Studies, Competitive Enterprise Institute
The Competitive Enterprise Institute (CEI) is a non-profit public
policy research foundation dedicated to individual liberty, limited
government, and markets. We appreciate the opportunity to discuss
policy issues surrounding online advertising.
Privacy dilemmas are inevitable on the frontiers of an evolving
information era, but CEI maintains that competitive approaches to
online privacy and security will be more nimble and effective than
rigid political mandates at safeguarding and enhancing consumer well-
being, facilitating commerce and wealth creation, and even contributing
to the rise of the anonymous approaches to commerce we'd like to see.
The Rise of Privacy and Cybersecurity as Public Policy Issues
The marvelous thing about the Internet is that one can contact and
learn about anyone and anything. The downside is that the reverse is
often true. The digital information age--against a backdrop of rising
globalization--offers consumers unprecedented access to news,
information, democratized credit and much more. Anyone may collect and
share information on any subject, corporation, government--or in many
cases, other individuals.
Companies from retailers to search engines to software makers all
collect consumer data--enough to fill vast server warehouses. Of
course, websites have long collected and marketed information about
visitors. The latest twist is that behavioral marketing firms ``watch''
our clickstreams to develop profiles or inform categories to better
target future advertisements. Unarguably beneficial, the process stokes
privacy concerns. Fears abound over the data's security; is any of it
personally identifiable? If not, can it conceivably become so? Will
personal information fall into the wrong hands? Will it become public?
And if a breach occurs, who's punished? While Capitol Hill, beltway
regulators or state governments are seen often as the first line of
defense, regulatory and legislative proposals, much like the anti-spam
law, can fall short of success. Aspirations can exceed actual
legislative capability.
Clearly, as a technological phenomenon, mass transactional data
tracking and collection are here to stay; and with nascent technologies
like biometrics that could fully authenticate users on the horizon, the
debates will only intensify.
Along with behavioral advertising, new data-mining and biometrics
technologies promise higher levels of convenience and, ultimately, more
secure commerce online. Beyond the ``merely'' commercial, the
technologies also hint at greater physical security in the ``homeland''
and in our workplaces via authentication.
On the upside, online advertising enables today's familiar
subscription-fee-free cornucopia of news and information, and the free
soapbox enjoyed by bloggers worldwide. It's become cliche to note the
commercialized Internet is one of the most important wealth-creating
sectors and democratizing technologies ever known. Benefits to society
range from frictionless e-commerce, to the democratization of
privileges once available only to the rich, to a megaphone for all.
This online bounty has also brought real and imagined privacy
vulnerabilities to the forefront, ranging from personal identity theft
to exposure of private thoughts and behavior online. Once, we could
contend merely with nuisances like spam, cookie-collection practices
and the occasional spyware eruption. Since policies today are being
formulated in the context of a post-Sept. 11 world, cybersecurity and
computerized infrastructure access and security join routine privacy as
prime policy issues. Adding complexity is the noted emergence of
biometric technologies and highly engineered data mining that could
alter the future of behavioral marketing. Thus we must contend not just
with run of the mill commercial aspects of privacy policies, but with
national security themes and what some consider a dangerous new
surveillance state.
The question is, do newfangled data collection techniques threaten
fundamental expectations of privacy, and in the case of government data
collection, even liberty itself?
What principles distinguish between proper and improper uses of
personal information, and what policies maximize beneficial e-commerce
and consumer welfare? Business use of behavioral advertising can be
irritating, but many have made peace with advertisers' using personal
information. One-size-fits-all privacy mandates will undermine e-
commerce and the consumer benefits we take for granted. Sweeping
regulations can especially harm start-ups that lack the vast data
repositories already amassed by their larger competitors. Our policies
should be consistent with tomorrow's entrepreneurs (and consumers)
starting businesses of their own to compete with the giants of today.
Thus, privacy policies need to be filtered through the lens of the
entire society's needs. We must consider the impact on: (1) consumers,
(2) e-commerce and commerce generally, (3) broader security,
cybersecurity, homeland security and critical infrastructure issues,
and finally (4) citizen's 4th amendment protections.
Happily the prospect of billions in economic losses from mistakes
incentivize the market's efforts to please consumers and safeguard
information and networks.
Web Functionality Continues to Unfold
The recent emergence of behavioral advertising reinforces the
easily forgotten reality that there's more to the Internet than the
``Web'' at any given juncture; it's only 2008, and there are doubtless
more commercially valuable avenues for marketing yet to be discovered
in the decades ahead. Targeted, behavioral and contextual advertising
make use of heretofore unexploited underlying capabilities of the
Internet, possibilities that hadn't yet occurred to anyone else, just
as the original banner ad trailblazers first did years ago--and, yes,
just as the spammers did.
At the Outset: Policy Must Distinguish Between Public and Private Data
Parameters are needed to talk coherently about the treatment of
individual's data. Information acquired through the commercial process
must be kept separate from that extracted through government mandates.
Similarly, private companies generally should not have access to
information that government has forced individuals to relinquish (what
one might call the ``Social Security'' problem). Private industry
should generate its own marketing-related information (whether
``personally identifiable'' or not), for purposes limited by consumer
acceptance or rejection, rather than piggyback on government IDs.
Confidentiality is a value, and should be a competitive feature.
Conversely, for any debate over behavioral advertising to make
sense, corporate America needs to be able to make credible privacy
assurances to the public. People need to know that the data they
relinquish is confined to an agreed-upon business, transactional or
record-keeping purpose, not incorporated in a government database. If
regulators end up routinely requiring banks, airlines, hotels, search
engines, software companies, Internet service providers and other
businesses to hand over private information (in potentially vulnerable
formats), they will not only undermine evolving commercial privacy
standards, including behavioral, but make them impossible. Government's
own information security practices is the elephant in the room when it
comes to contemplating e-commerce sector's stance with respect to
privacy. It's all too easy to give the online marketing industries a
black eye and risk turning society against the technologies, and ensure
regulation and politicization. Private data and public data policies
are potentially on a collision course, but need not be.
The benefits that personalization brings, like easier, faster
shopping experiences, are in their infancy. Sensible data collection
improves search, communication, ability to innovate, U.S.
competitiveness--all the things we associate with a well-functioning
economy and evolution in healthy consumer convenience and power.
Privacy Legislation: Premature and Overly Complex
In contemplating government's role with respect to privacy and
information security, we must recognize the realities of differing user
preferences that preclude one-size-fits-all privacy and security
policy. Online, there are exhibitionists and hermits. Some hide behind
the equivalent of gated communities; others parade less-than-fully
clothed before personal webcams.
Note how we work ourselves up into a lather: policymakers were
concerned about privacy when ads were untargeted and irrelevant (spam);
now a solution--behavioral and contextual marketing--makes ads
relevant, and we're hand-wringing about privacy there too.
Incidentally, spam was framed as a privacy problem, but in reality the
spammer didn't typically know who you were. Likewise, a positive early
development in behavioral advertising is that personally identifiable
information is not always crucial to the marketer (although sensible
uses of personally identifiable information should not be thwarted).
Too often, the complaint seems to be commerce as such. For example, the
Federal Communications Commission recently decided to investigate the
``problem'' with embedded ads in TV programming.\1\
---------------------------------------------------------------------------
\1\ Associated Press, ``FCC to look into embedded advertising on
TV,'' MSNBC.com. June 26, 2008. http://www.msnbc.msn.com/id/25401193/.
---------------------------------------------------------------------------
Policy should recognize privacy is not a single ``thing'' for
government to protect; it is a relationship expressed in countless
ways. That relationship is best facilitated by emergent standards and
contracts--like the Network Advertising Initiative's behavioral
advertising principles \2\ that predate the Federal Trade Commission's
late 2007 principles \3\--and in emergent market institutions like
identity theft insurance. Apart from varied privacy preferences, any
legislative effort to regulate behavioral advertising gets exceedingly
complex:
---------------------------------------------------------------------------
\2\ http://www.networkadvertising.org/networks/
principles_comments.asp.
\3\ Federal Trade Commission, ``Behavioral Advertising, Moving the
Discussion Forward to Possible Self-Regulatory Principles,'' December
20, 2007. http://www.ftc.gov/os/2007/12/P859900stmt.pdf.
---------------------------------------------------------------------------
If online privacy is regulated, what about offline?
Should behavioral advertising be opt-in or opt-out? (Why and
when?)
Who defines which advertising is ``behavioral''?
What is the legislative line between sensitive, and non-
sensitive, personally identifiable information?
Should the Federal Government pre-empt state privacy laws?
Will the privacy rules apply to government?
Will government abstain from accessing or seizing private
databases?
What about non-commercial information collection? (Will the
rules apply to bloggers? Or to Facebook activism?)
What about consumer harm caused by privacy legislation
(Given that in the business world, most transactions occur
between strangers.)
What of practical problems of written privacy notices?
(Especially given the declining importance of the desktop, the
emergent web-like multi-sourced nature of web-pages themselves,
smaller wireless-device screens, and the ``thing-to-thing'' Net
that bypasses humans altogether.)
Could disclosure and reporting mandates create a burdensome
paperwork requirements detrimental to small businesses? (A
privacy ``Sarbanes-Oxley'')
What about the right to remain anonymous; Behavioral
marketing appears to be on course to facilitate anonymous
transactions; will government permit it? How should tolerance
of anonymity differ in commercial and political contexts?
The Internet was designed as an open, non-secure network of semi-
trusted users. Thus one interpretation of the nature of the cyberspace
is that advertisers may legitimately assemble information on what is
clearly a very public network that never offered any real pretense of
security. But even assuming one's online pursuits can be tracked,
privacy tools nonetheless are emerging, and vendors must be held to
commitments. Given legislation's complications and the Internet's
inherent security limitations, a rational policy prescription should be
more limited: Hold the private sector accountable to the contracts and
guarantees it makes, and target identity theft and the criminals who
perpetrate it. If legislation merely does such things as send bad
actors overseas, we merely create regulatory hassles for mainstream
companies that already follow ``best practices,'' and for small
businesses trying to make a go of legitimate e-commerce.
As in spam debate, we face less a legislative problem than a
technological one. It's true that social norms and expectations have
yet to gel--but those are as varied as individuals are.
Marketing Is Not Today's Dominant Information Collection Threat
The emphasis on online privacy legislation could represent a case
of misdirected energy. The most important information collection issues
of the day are not related to mere marketing; rather, criminals who
ignore already existing laws and will ignore any new law, are the ones
creating mischief online, abusing the trust we have or would like to
have in vendors. Meanwhile, government surveillance and information
collection threaten liberties and genuine privacy--and one cannot ``opt
out.'' (One is reminded of the Peanuts cartoon of Snoopy sitting on his
doghouse typing, ``Dear IRS . . . Please remove my name from your
mailing list.'') \4\
---------------------------------------------------------------------------
\4\ http://www.freerepublic.com/focus/f-news/1384722/posts.
---------------------------------------------------------------------------
The stringent opt-in standard some seek in the behavioral marketing
debate is not one government tolerates for itself. The post-Sept. 11
push for compulsory national ID cards, warrant-less wiretapping and
escalating data retention mandates signify a government more inclined
toward infringing privacy than acting as guarantor.
The rise of the information society amid a ``homeland security
culture'' is an unfortunate coincidence, an accident, but one that
colors debates over marketing that would otherwise be more pedestrian.
The tendency of government to interfere with privacy practices is
undeniable: Total Information Awareness, CAPPSII, and a national ID are
examples of expansive government efforts that would undermine the
private sector's freedom and ability to make privacy assurances in the
first place.
Worse, when technology companies contract with government for
information services, they would very likely request immunity for data
breaches by extension of the Homeland Security Act that grants similar
immunities for failed security technologies; so if markets are tempted
to repudiate self-regulation and liability for privacy standards,
government oversight becomes the default. The ``homeland security
culture'' can undermine the market's entrepreneurial tendency to
resolve the dilemmas created by information sharing.
Deliberations over privacy and online security should start with
the recognition that government often doesn't need to protect our
privacy, it needs to allow it in the first place. Business, whatever
missteps happen in behavioral marketing, can deliver. As it stands,
nobody's in any position to make ironclad security guarantees given the
open nature of the Internet, but the Web is a giant research
experiment, and techniques will improve. In fact, as behavioral
tracking does begin to employ personally identifiable information,
security benefits in ways that people will approve. The Net's
governmental origins have left privacy expectations and rights somewhat
ill-defined in many online contexts. But we all at times need to
identify ourselves and validate the identity of others.
Consumers are Not Powerless: The Redundancy of FTC Standards
In spite the Net's vulnerabilities, consider how legislation pales
compared to unforgiving competitive discipline. An old joke holds that
if McDonald's was giving away free Big Macs in exchange for a DNA
sample, there would be lines around the block. But consumers do care;
and thanks to the Internet itself, they are hardly a voiceless mass.
Every few weeks brings new headlines about government data-handling
debacles, such as governmental bodies forcing employees to carry Social
Security cards on their person, or the IRS requirement that payment
checks feature the SSN.\5\ Confidence isn't inspired when the
government's information practices lag the private sector's.
---------------------------------------------------------------------------
\5\ Associated Press, ``U.S. Contradicts Itself Over Its Own ID
Protection Advice,'' SiliconValley.com, July 2, 2008. http://
www.siliconvalley.com/news/ci_9762027?nclick
_check=1.
---------------------------------------------------------------------------
Contrast that with what happens to a careless private firm. Google
and its recent mergers and alliances put it under scrutiny, but why?
(Recall it was Google that in 2006 refused to hand over user search
data to the Justice Department; and Google's YouTube division is now
being forced by a New York district court to hand over user viewing
records in a video piracy case. Google not unsurprisingly objects.) But
imagine if Google suffered a serious data breach. Consumers would lose
trust, and Google could lose millions. Examples abound of consumer
sovereignty, such as the backlash against Facebook's Beacon that cross-
posted users shopping activities on friends' sites,\6\ and Comcast's
de-prioritizing of certain file sharing transfers. Today's Internet
users are empowered to educate the world about business practices of
which they disapprove. The blogosphere transforms Web users into
citizen-journalists, harnessing the power of collective discontent. The
result: Companies routinely change and improve their information
handling procedures without law.
---------------------------------------------------------------------------
\6\ Caroline McCarthy, ``MoveOn.org takes on Facebook's `Beacon'
ads,'' CNet News.com. November 20, 2007. http://news.cnet.com/8301-
13577_3-9821170-36.html.
---------------------------------------------------------------------------
Policies proposed in the name of what consumers want or should want
are all too common, as if the ideas hadn't occurred to anyone in the
competitive marketplace already, or as if the markets hadn't been
forced to adapt already, or as if issues weren't more complicated than
the regulators suppose.
For example, the November 2007 FTC proposal on behavioral
advertising offers pedestrian principles that have long been in play:
\7\ Paraphrasing, sites should declare that info is being collected and
used and users can opt out; data should be ``reasonably secured,'' and
retained only as long as necessary; affirmative consent be given for
privacy policy changes; and sensitive information should not be
collected at all, or only with affirmative opt-in.
---------------------------------------------------------------------------
\7\ Federal Trade Commission, 2007.
---------------------------------------------------------------------------
Where do the real incentives lie? Industry looks at what consumers
actually want; industry often already embraces opt-in for sensitive
information categories, even when the information is not personally
identifiable. And if not so empowered by a benevolent vendor, users can
already exercise the choice allegedly sought in privacy legislation;
they can simply choose not to disclose sensitive information on certain
sites, or employ privacy software that can thwart unwanted data
collection and allow anonymous Web browsing. ``Anonymizer'' is still
out there for encrypted, anonymous surfing. People can switch to
``Scroogle'' to disguise their Google searches; A consumer can use a
dedicated tool to nullify his identity prior to a sensitive search like
``HIV''; TrackMeNot can send out ``white noise'' search queries to
disguise the real one. No mandates for choice are needed; choice is the
default, whether vendors prefer it or not.
In terms of competitive enterprise, the divisiveness of a debate
like behavioral marketing implies that real market opportunities exist
in providing online anonymity. After all, despite all the hand-wringing
over personally identifiable information, any given marketer doesn't
necessarily need to know who you are, but how somebody like you acts.
(Much like a politician seeking a vote, incidentally.) Again, the worry
is less that the market is invading our privacy and more whether that
anonymity will be permitted politically when it finally is available to
us commercially.
``Self-Regulation'' Is a Misnomer
Privacy and security need to be competitive features. We need to
foster competition in reputations. And we need flexibility when the
inevitable mistakes are made.
Businesses compete; and one area in which they can compete is in
the development of technologies that enhance security. Washington's
inclination toward regulating online consumer relationships threatens
to undermine the market's catering to diverse individual privacy
preferences, and hinder the evolution of competitive research and
innovation in secure applications. Privacy encompasses innumerable
relationships between consumers and businesses, and no single set of
privacy safeguards is appropriate. While government demands information
disclosure, profit-driven firms compete to offer robust privacy
assurances. As businesses respond to evolving consumer preferences,
stronger privacy policies will emerge.
Businesses are disciplined by responses of their competitors.
Political regulation is pre-mature; but ``self-regulation'' like that
described in the FTC principles is a misnomer; it is competitive
discipline that market processes impose on vendors. Nobody in a free
market is so fortunate as to be able to ``self regulate.'' Apart from
the consumer rejection just noted, firms are regulated by the
competitive threats posed by rivals, by Wall Street and intolerant
investors, indeed by computer science itself.
Neither the government nor private sector has a spotless ``self-
regulatory'' record, but FTC seems unconcerned about the former. Data
breaches at businesses, governments and universities rose 69 percent in
2008.\8\ Government can contribute to data security by ensuring that
its own policies--like data sharing or data retention mandates, or
sweeping subpoenas--do not interfere with competitive discipline.
---------------------------------------------------------------------------
\8\ Brian Krebs, ``Data Breaches Are Up 69% This Year, Nonprofit
Says,'' Washington Post. July 1, 2008. p. D3. http://
www.washingtonpost.com/wp-dyn/content/article/2008/06/30/
AR2008063002123.html.
---------------------------------------------------------------------------
Even governmental calls for self-regulation seem lukewarm. Along
with the Federal Trade Commission's Principles on what personally
identifiable information firms may collect, a bill in the New York
state legislature would impose drastic opt-in standards, preventing
companies from gathering personalized information without explicit user
permission. When Microsoft bid for Yahoo! this year, the Justice
Department almost immediately wondered whether the combined firm would
possess ``too much'' consumer data. Canada recently announced an
investigation into Facebook's privacy protections. Now the Department
of Justice is investigating the Google-Yahoo deal.\9\
---------------------------------------------------------------------------
\9\ Peter Whoriskey, ``Google Ad Deal Is Under Scrutiny,''
Washington Post, July 2, 2008.
Page D1. http://www.washingtonpost.com/wp-dyn/content/article/2008/07/
01/AR200807010
2622.html.
---------------------------------------------------------------------------
Everybody's heard of Google and Microsoft, but fewer have heard of
companies like Phorm and NebuAd, which present the more pertinent
behavioral marketing issues; their new techniques give ISPs a dog in
the fight, since online advertising is a commercial opportunity
impossible for ISPs to ignore. ISPs see Google and Microsoft and they
want a piece of the online advertising action too. These companies'
techniques have been called spyware, but again, they incorporate the
Net's underlying capabilities in novel ways, and they too are subject
to competitive discipline. One's sympathies will depend upon the
``ownership'' status one accords to Web pages, and what one regards as
online ``trespass.'' The only certainty is a Web page today is not what
a Web page tomorrow will be. Was there ever a real reason for
publishers and advertisers to think they could control everything a
user saw, given the open-ended potential of software's obvious ability
to route content to browsers in novel ways? At many sites, like
Facebook, each page is a ``Web'' in its own right, containing widgets
drawing information and ads from numerous sources. The debate has
really only just begun, and online marketing trade groups are truly the
``Battered Business Bureau.'' But they're battered by competitive
discipline, not merely regulators.
Lessons from Personally Identifiable Data Use Can Inform Future Online
Security Practices
A frontier industry requires the flexibility to learn from
mistakes. We must distinguish between proper and improper uses of
surveillance by both the private and public sectors. Not many want to
be tracked by the authorities, or treated like human bar code. Myriad
benefits will accrue from the further deployment of identification
techniques--even personally identifiable--into various facets of daily
life. But where is the line crossed, and who is capable of crossing it?
In private hands, techniques like behavioral marketing, biometric
and data-mining technologies enlarge our horizons. They expand the
possibilities of a market economy by bolstering security in private
transactions ranging from face-to-face authentication to long-distance
commerce. The best, most secure technologies are those that prevent
others from posing as us--that's why the value of personally
identifiable data cannot be ruled out. The Web is desperately short of
that kind of clarity and authentication, in a world of cyber-risks,
identity theft, and the need to conduct ever more sensitive
transactions. But nothing is automatic. The marketplace imperative
requires private sector experimentation in privacy: It's messy, but
necessary.
On the one hand, policy should not create situations where
companies are required to ask for personal info that otherwise wouldn't
be needed. (Google declares in its comments on the FTC advertising
principles that obeying certain rules would require it to collect
information it otherwise would not need.) On the other hand, certain
forms of identifiable behavioral tracking may prove important in
specific contexts and shouldn't be prohibited.
Disallowing personally identifiable information is the wrong thing
to do. We often need to identify those we're dealing with on line, and
for them to be able to identify us; such instruments will be governed
by heretofore unknown contracts and privacy polices. It's not ``self-
regulation,'' but the needs of the world at large driving this
evolution. Rather than legislating, it's likely better to keep this a
war between computer scientists; between those working on behavioral
advertising with personal information and/or authentication, and those
working on behavioral without authentication. Being able to sell to a
customer but not have that customer identified is a key research area
in computer science. The consumer-control ethos--the notion that we
don't have to be tracked--puts consumers, not advertisers, in the
drivers' seat Let the computer scientists duke it out.
In many transactions and contexts, the Web needs better
authentication, not the abandonment of personally identifiable
information. The private sector should experiment with generating such
data in ways that consumers can accept. Some say we must regulate
because online risks exist; this report argues for not regulating
because there are online risks. The firms that reduce risks in ways
palatable to consumers offer a great service. New products and
institutions still need to emerge around online commerce.
Expanding the Marketplace for Liability and Private Security Insurance
Privacy is one subset of the much broader issues of online security
and cybersecurity. It's been noted that a basic problem today is that
no one stands in any position to make guarantees to anybody about
anything. That doesn't mean improved insurance products and enhanced
liability contracts won't develop online, however. Lessons learned from
spam, privacy, and preventing piracy of intellectual property will
carry over to the security issues of tomorrow.
Government shouldn't grant immunity to software companies for
breaches, but at the same time it should not impose liability on them
either. It's not so clear whom to sue on an Internet not amenable to
authentication, but standards will emerge. Government interference can
impede private cyber-insurance innovations.
Certain innovations can be sacrificed by regulating. The private
sector needs to ``practice'' now for the really difficult cases like
the integration of biometrics into the online world; meanwhile the
Federal Government needs to focus on cyber-crime.
A Positive Agenda for the Federal Government
Policymakers should appreciate the government's inherent
limitations as well as the vulnerabilities that can be created by
Federal policies and procedures.
From lost laptops to hacks into the Pentagon e-mail system, to
``D'' grades for the Department of Homeland Security's own information
security practices, regulators' ability to rationally guide others on
privacy is questionable. In many areas it makes sense to circumscribe
regulators' sphere of influence, while increasing that of the market.
Recognizing that governments can fail just as markets can, there
are numerous ways government within its limitations can properly foster
private sector innovation in security:
Foster competitive discipline.
Emphasize protecting government's own insecure networks, not
regulating markets. This means many things, including: removing
sensitive information from government websites; limit the size
and scope of government databases to ensure government doesn't
create artificial cybersecurity risks; avoiding data retention
mandates and other interventions that undermine private-sector
security guarantees.
Focus on computer criminals, not cyber-regulations.
Assess areas where it's best to liberalize private sector
data-sharing rules. For example, facilitating private sector
medical data sharing could deliver benefits to suffering
patients. More broadly, some firms cannot share data among
their own divisions because of antitrust and privacy
strictures. Enhancing cross-firm coordination can improve
reliability and security.
Recognize that commercial anonymity and political anonymity
differ; we may need ``less'' of the former, even as we expand
the latter. Research should continue on the seemingly opposed
agendas of authentication of users on the one hand, and
anonymizing technologies on the other.
Conclusion: Affirming Private Sector Primacy Over Information Practices
Our greatest privacy concern should be government collection of our
information, not the emergence of targeted marketing.
In the changing world of e-commerce, the role of government is not
to predetermine commercial privacy arrangements, but to enforce
information-sharing contracts that companies make between themselves or
with individuals. Privacy policies are legally binding. Government's
role is not to dictate the structure of privacy contracts through such
means as opt-in or opt-out policies; it is to halt deceptive practices
and hold private firms accountable to the guarantees they make.
Government's other role is to protect citizens from identity theft,
which is not a commercial enterprise, but a criminal one.
If anonymity and the inability to exclude bad actors are at the
root of genuine online security problems, legislation doesn't make them
go away. When contemplating centralized government vs. decentralized
market approaches to protection consumers online, we must strive,
before regulating, to follow the ``cybersecurity commandment'': Don't
entrench regulation to such a degree that effective private
alternatives and institutions, however warranted as conditions change,
simply cannot emerge.
Related Reading
Wayne Crews and Ryan Radia, ``Rigid Federal Mandates Hinder Privacy
Technologies,'' San Jose Mercury News, June 15, 2008, http://
www.mercurynews.com/opinion/ci_9593341.
Wayne Crews, ``Cybersecurity Finger-pointing: Regulation vs.
Markets for Software Liability, Information Security, and Insurance,''
CEI Issue Analysis 2005 No. 7, May 31, 2005,
http://cei.org/pdf/4569.pdf.
Wayne Crews, ``Cybersecurity and Authentication: The Marketplace
Role in Rethinking Anonymity--Before Regulators Intervene,'' CEI Issue
Analysis 2004 No. 2, November 8, 2004, http://cei.org/pdf/4281.pdf.
Wayne Crews, Comments to the FTC on e-mail authentication themes,
September 30, 2004, http://www.cei.org/pdf/4229.pdf.
Alberto Mingardi and Wayne Crews, EU takes a Swipe at Google,
International Herald Tribune, March 9, 2007, http://www.iht.com/
articles/2007/03/09/opinion/edmingardi.php.
Wayne Crews and Brooke Oberwetter, ``Preventing Identity Theft and
Data Security Breaches: The Problem With Regulation,'' CEI Issue
Analysis 2006 No. 2, May 9, 2006, http://cei.org/pdf/5316.pdf.
Wayne Crews ``Giving Chase in Cyberspace: Does Vigilantism Against
Hackers and File-sharers Make Sense?'' CEI OnPoint No. 109, October 2,
2006. http://cei.org/pdf/5569.pdf.
Wayne Crews, ``Trespass in Cyberspace: Whose Ether Is It Anyway?''
TechKnowledge #19, Cato Institute, September 10, 2001, http://
www.cato.org/tech/tk/010910-tk.html.
Wayne Crews, ``Human Bar Code: Monitoring Biometrics Technologies
In a Free Society,'' Cato Institute Policy Analysis No. 452, September
17, 2002, http://www.cato.org/pubs/pas/pa452.pdf.
Senator Dorgan. Mr. Crews, thank you very much.
Finally, we will hear from Mr. Mike Hintze.
Mr. Hintze. Hintze.
Senator Dorgan. Mr. Mike Hintze. I am sorry. Mike Hintze is
the Associate General Counsel at Microsoft Corporation. Mr.
Hintze, you may proceed.
STATEMENT OF MICHAEL D. HINTZE, ASSOCIATE GENERAL COUNSEL,
MICROSOFT CORPORATION
Mr. Hintze. Thank you, Mr. Chairman, for inviting me to
testify today about the privacy implications of online
advertising. This is a critically important topic that
Microsoft takes very seriously, and we applaud the Committee
for its leadership in this area.
Online advertising, as you acknowledged, has become the
engine that drives the Internet economy. Millions of websites
are able to offer content and services for free to consumers
because of the revenue they derive from advertising online. In
the United States, the amount spent on online advertising
already exceeds spending for advertising through radio,
magazines, and cable television. It accounted for $21 billion
in 2007 and is expected to grow to $50 billion in the next 3
years.
Online advertising has been so successful because it is
interactive and can be targeted to users' online activities and
other characteristics. This targeting benefits users not only
because it enables free services and content they enjoy, but
also because they are likely to see more relevant ads. And it
benefits advertisers because they can reach users who are more
likely to respond to their ads.
Each search, click, and other user action online reveals
valuable information about the user's likely interests, and
online ads can be automatically tailored to those interests. In
general, most data collection online happens in conjunction
with a display of ads. This means the entity with the greatest
market share and therefore who serves the most ads online will
collect the most data about users.
As this Committee recognizes, the collection of user data
to serve ads on the Internet has important privacy
implications. Microsoft is here today because we have a deep
commitment to consumer privacy. We were one of the first
companies to appoint a chief privacy officer, an action we took
nearly a decade ago, and we currently have over 40 employees
who focus on privacy full-time and another 400 throughout the
business who focus on privacy as part of their jobs. We have a
robust set of internal privacy policies and standards that
guide how we do business and how we design our products and
services in a way that protects consumer privacy.
With respect to online advertising, we have taken more
concrete steps to protect privacy than any of our competitors.
Last July, we released Microsoft's Privacy Principles for Live
Search and Online Ad Targeting. We are committed to these
principles which focus on three core themes: transparency,
control, and security. Let me explain how we have put each of
these principles into action in ways that go beyond others in
the industry.
The first principle is transparency. We post a clear link
to our privacy notice on every page of our websites, including
the home page, and we have for several years. We also were one
of the first companies to develop so-called layered privacy
notice, which gives concise and easy-to-understand bullet-point
summaries of our practices with links to more detailed
information. And our privacy statement is clear about the data
we collect and use for online advertising.
The second principle is control. Microsoft enables users to
opt out of behavioral ad targeting, but we also give consumers
the option to tie their opt-out choice to their Windows online
account. Unlike methods used by other companies, this means
that even if they delete cookies on their machine, when they
sign back in, their opt-out selection will persist. It also
means that a single choice can apply across multiple computers
they use.
The third principle is security. For Microsoft, this means
not only protecting data from unauthorized access, but also
designing our systems and processes in ways that minimize their
privacy impact from the outset. We create an anonymized
identifier for each of our registered users. Search query data
and web surfing behavior used for ad targeting is associated
with this anonymized identifier. We also irreversibly and
completely remove the IP addresses and other identifiers from
search queries after 18 months.
We believe that our commitment to these three principles,
transparency, control, and security, and more importantly, the
steps we have taken to implement them make us the industry
leader in online privacy.
These principles also form the basis for our support for
comprehensive baseline privacy legislation, supplemented by
robust self-regulation. For example, we have advocated a
broader self-regulatory framework than that proposed by the
FTC, one that is tailored to account for the type of
information collected and how it is used. We have also long
supported meaningful privacy legislation which, as CDT
appropriately notes, can protect consumers without hampering
business.
We view these efforts as part of our multi-faceted approach
to protecting consumer privacy, which also includes developing
technical solutions and educating consumers about how to
protect themselves online. In short, at Microsoft, we are
prepared to work collaboratively on all these fronts to protect
consumer privacy.
Thank you for giving us the opportunity to testify today,
and I look forward to answering any questions you may have.
[The prepared statement of Mr. Hintze follows:]
Prepared Statement of Michael D. Hintze, Associate General Counsel,
Microsoft Corporation
Chairman Inouye, Vice Chairman Stevens, and honorable Members of
the Committee, my name is Michael Hintze, and I am an Associate General
Counsel of Microsoft Corporation. Thank you for the opportunity to
share Microsoft's views on the important privacy issues presented by
advertising on the Internet. We appreciate the initiative that this
Committee has taken in holding this hearing, and we are committed to
working collaboratively with you, the Federal Trade Commission,
consumer groups, and other stakeholders to protect consumers' privacy
interests online.
Much is at stake with respect to the issues we will be considering
today. Online advertising has become the very fuel that powers the
Internet and drives the digital economy. It supports the ability of
websites to offer their content and services online; it has created new
opportunities for businesses to inform consumers about their products
and services; and it allows consumers to receive ads they are more
likely to find relevant. Simply stated, the Internet would not be the
diverse and useful medium it has become without online advertising.
At the same time, online advertising is unique because it can be
tailored automatically to a computer user's online activities and
interests. An online ad can be served based on the website a user is
visiting, the searches a user is conducting, or a user's past Internet
browsing behavior, among other things. In each instance, serving the
online advertisement involves the collection of information about
consumers' Internet interactions. And this data collection has
implications for consumer privacy.
The objective we face is to maintain the growth of online
advertising while protecting consumer privacy. This is a commitment
Microsoft embraces. We recognize that consumers have high expectations
about how we and other Internet companies collect, use, and store their
information. Consumers must trust that their privacy will be protected.
If the Internet industry fails to meet that standard, consumers will
make less use of online technologies, which will hurt them and industry
alike.
It also could hurt the U.S. economy. E-commerce sales reached
$136.4 billion in 2007, an increase of 19 percent from 2006, according
to the U.S. Census Bureau.\1\ In comparison, total retail sales in 2007
increased only 4 percent from 2006. If consumers feel that Internet
companies are not protecting their privacy, the Internet's ability to
serve as an engine of economic growth will be threatened. This means
that Microsoft, and all companies operating online, must adopt robust
privacy practices that build trust with consumers.
---------------------------------------------------------------------------
\1\ U.S. Census Bureau, Quarterly Retail E-Commerce Sales: 4th
Quarter 2007, Feb. 15, 2008, available at http://www.census.gov/mrts/
www/data/html/07Q4.html.
---------------------------------------------------------------------------
Microsoft has a deep and long-standing commitment to consumer
privacy. Microsoft was one of the first companies to appoint a chief
privacy officer, an action we took nearly a decade ago, and we
currently employ over 40 employees who focus on privacy full-time, and
another 400 who focus on it as part of their jobs. We have a robust set
of internal policies and standards that guide how we do business and
how we design our products and services in a way that respects and
protects user privacy.\2\ And we have made significant investments in
privacy in terms of training and by building our privacy standards into
our product development and other business processes.
---------------------------------------------------------------------------
\2\ Some of these standards are set forth in Microsoft's Privacy
Principles for Live Search and Online Ad Targeting, attached as
Appendix 1.* This document is also available at http://
www.microsoft.com/privacy. Additionally, Microsoft's Privacy Guidelines
for Developing Software Products and Services, which are based on our
internal privacy standards, are available at http://www.microsoft.com/
privacy.
---------------------------------------------------------------------------
In general, three key principles have guided our approach to
privacy issues:
Transparency. We believe consumers should be able to easily
understand what information will be collected about them and
when. They also should know how such information will be used
and whether it will be combined with other information
collected from or about them.
Control. We believe consumers should be able to control
whether their personal information is made available to others
and should have a choice about whether information about their
online activities is used to create profiles for targeted
advertising.
Security. Consumers and their information should be
protected against outside threats and from unwanted disclosure.
Data that directly identifies individual consumers, such as
name and e-mail address, should not be stored in direct
association with search terms or data about Web surfing
behavior used to deliver ads online. And strict data retention
policies should apply to search data.
Today, I will discuss why we believe these principles are
important, how we have put each of these principles into action, and
how they underlie Microsoft's approach to privacy in online
advertising. But first I would like to provide an overview of how
online advertising works, the role that consumer data plays in serving
online ads, and the online advertising market.
I. Online Advertising and the Role of User Data
Consumers today are able to access a wealth of information and a
growing array of services online for free. Websites can offer this
content and these services for free because of the income they receive
from advertising.\3\ Just as newspapers and TV news programs rely on
traditional advertising, online news sites and other commercial
websites rely on online advertising for their economic survival. Online
advertising is particularly critical for the thousands of smaller
websites that do not publish through offline channels and thus depend
entirely on the revenue they receive from selling space on their
websites to serve ads online. It is also critical for smaller
businesses that serve niche markets (e.g., out-of-print books on
European history) who rely on online advertising to reach those niche
audiences cost-effectively; indeed, many of these businesses could not
survive without it.
---------------------------------------------------------------------------
\3\ It has become a standard approach to the online economy that
there is a value exchange in which companies provide online content and
services to consumers without charging a fee and, in return, consumers
see advertisements that may be targeted.
---------------------------------------------------------------------------
The importance of online advertising is evident from its growing
share of the overall advertising market. It accounted for $21 billion
of the market in 2007 and is expected to grow to $50 billion in the
next 3 years.\4\ In the United States, online advertising spending
already exceeds spending for advertising through radio, magazines, and
cable television.\5\
---------------------------------------------------------------------------
\4\ See Interactive Advertising Bureau, IAB Internet Advertising
Revenue Report, 7, May 2008, available at http://www.iab.net/media/
file/IAB_PwC_2007_full_year.pdf; Yankee Group, Yankee Group Forecasts
U.S. Online Advertising Market to Reach $50 Billion By 2011, Jan. 18,
2008, available at http://www.yankeegroup.com/
pressReleaseDetail.do?actionType=getDetail
PressRelease&ID=1805.
\5\ See Brian Morrissey, IAB: Web Ad Spend Tops Cable, Radio,
Adweek, May 15, 2008, available at http://www.adweek.com/aw/
content_display/news/digital/e3ibcf6d45fc7a036dff28457
a85c838ff1.
---------------------------------------------------------------------------
One reason for this rapid growth is the ability to target online
ads to Internet users. Newspaper, magazine, and television
advertisements can, of course, be targeted based on the broad
demographics of readers or viewers. But the Internet is interactive,
and this interaction yields a wealth of data about users' activities
and preferences. Each search, click, and other user action reveals
valuable information about that user's likely interests. The more
information an entity collects, the greater that entity's ability to
serve an advertisement that is targeted to the user's interests. This
targeting benefits users, not only because it enables the free services
and content they enjoy, but also because the ads they see are more
likely to be relevant. And it benefits advertisers because users are
more likely to respond to their ads.\6\
---------------------------------------------------------------------------
\6\ It is for this reason advertisers are willing to pay more for
targeted ads. For example, although Merrill Lynch has reported that the
average cost per 1,000 impressions (``CPM'') is $2.50, entities engaged
in behavioral targeting have reported average CPMs as high as $10. See
Brian Morrissey, Aim High: Ad Targeting Moves to the Next Level,
Adweek, Jan. 21, 2008, available at http://www.adweek.com/aw/magazine/
article_display.jsp?vnu_content_id=1003695822. Data also shows that 57
percent of 867 search engine advertisers and search engine marketing
agencies polled ``were willing to spend more on demographic targeting,
such as age and gender.'' Search Engine Marketing Professional
Organization, Online Advertisers Are Bullish on Behavioral Targeting,
May 15, 2008, available at http://www.sempo.org/news/releases/05-15-08.
---------------------------------------------------------------------------
There are a variety of ways in which data can be collected about
users to serve targeted ads on the Internet. Users reveal information
about what they are looking for when they search online, and ads can be
targeted to their search queries.\7\ Advertising networks enter into
agreements with websites that allow them to display ads; to deliver and
target those ads, data is gathered about the pages users view and the
links users click on within those sites.\8\ And new business models are
emerging where data about users' online activities can be collected
through a user's Internet service provider, and ads can be served based
on that information. In general, most data collection happens in
connection with the display of ads. This means the entity that serves
the most ads (search and/or non-search ads) will also collect the most
data about users.
---------------------------------------------------------------------------
\7\ Search ads are selected based on the search term entered by a
user and sometimes on data that has been collected about the user, such
as the user's history of prior searches. Search ads generally appear
either at the top of the search results or along the right-hand side of
the page. They often are displayed as text, but they may include
graphics as well. Advertisers bid against each other for the right to
have their ads appear when a specific search term is entered (known as
a ``keyword'').
\8\ These non-search ads are what users see when they visit
virtually any site on the Internet other than a search engine site.
They can be based on the content of the page the user is viewing
(typically referred to as ``contextual'' ads) or on a profile of a
user's activities that has been collected over time (referred to as
``behavioral'' ads). But in either case, the company serving the ad
would log the pages users view--typically in association with a cookie
ID from the user's computer and/or an IP address.
---------------------------------------------------------------------------
II. The Online Advertising Environment
The online advertising ecosystem has undergone significant changes
in the past few years. There continue to be millions of websites that
display online ads and thousands of advertisers who use online
advertising. However, there is a relatively small number of so-called
advertising networks, or ``middlemen,'' to bring advertisers and
websites together to buy and sell online ad space. And the number of
companies playing this intermediary role has decreased significantly in
recent months as a result of consolidation in the industry.\9\
---------------------------------------------------------------------------
\9\ Three examples of this are Microsoft's acquisition of
aQuantive, Yahoo!'s acquisition of RightMedia and Google's acquisition
of DoubleClick. For more information about the key players in the
advertising market and the impact of consolidation in the market, see
the testimony of Microsoft General Counsel Brad Smith before the Senate
Judiciary Committee, available at http://www.microsoft.com/presspass/
exec/bradsmith/09-27googledoubleclick.mspx.
---------------------------------------------------------------------------
This market consolidation impacts the privacy issues we are
discussing today in several ways. First, it is important to recognize
that in the past, advertising networks typically did not have direct
relationships with consumers. Today, however, the major ad networks are
owned by entities--such as Microsoft, Google, and Yahoo!--that provide
a wide array of Web-based services and, therefore, often have direct
relationships with consumers. This increases the potential that data
collected through online advertising will be combined with personally
identifiable information. While Microsoft has designed its online
advertising system to address this concern,\10\ no ad network is
required to do so.
---------------------------------------------------------------------------
\10\ See section III.C below.
---------------------------------------------------------------------------
Further, as noted above, there is a direct connection between the
market share of an advertising network or an online search provider and
the amount of data collected about a user's online activity. For
example, the larger the share of search ads a company delivers, the
larger number of users' online search queries it collects and stores.
Similarly, the larger the share of non-search ads an advertising
network delivers across the Web, the larger number of users' page views
it collects and stores, and the more complete picture of individuals'
online surfing behavior it is able to amass. Today, Google AdWords is
the leading seller of search advertising.\11\ Google also has the
leading non-search ad network, AdSense. Google recently expanded its
reach into non-search by acquiring DoubleClick.\12\ By comparison,
Microsoft is a relatively small player in search ads, and its reach in
non-search advertising is also smaller than Google's.\13\ Google's
growing dominance in serving online ads means it has access to and
collects an unparalleled amount of data about people's online
behavior.\14\
---------------------------------------------------------------------------
\11\Based on comScore's Core Search Report, in May of this year, 62
percent of searches were performed in the U.S. on Google, amounting to
roughly 6.7 billion searches. comScore, comScore Releases May 2008 U.S.
Search Engine Rankings, June 19, 2008, available at http://
www.comscore.com/press/release.asp?press=2275. Google also has
strategic agreements with AOL and Ask that allow Google to serve ads to
those companies' search engine sites. Adding AOL's (4.5 percent) and
Ask.com's (4.5 percent) share of the search queries, Google's share
rises to 71 percent. See id.
\12\ Following its acquisition of DoubleClick, Google now serves in
the range of 70 percent of all non-search advertisements. See, e.g.,
Lots of Reach in Ad . . ., April 1, 2008, available at http://
battellemedia.com/archives/004356.php.
\13\ Microsoft's Live Search has approximately 8.5 percent of Core
Search queries in the United States. comScore, comScore Releases May
2008 U.S. Search Engine Rankings, June 19, 2008, available at http://
www.comscore.com/press/release.asp?press=2275.
\14\ Concerns have been raised about this dominance as well as the
privacy protections surrounding the enormous amount of information
about users' online behavior that this dominance enables. See, e.g.,
Electronic Privacy Information Center, Supplemental Materials in
Support of Pending Complaint and Request for Injunction, Request for
Investigation and for Other Relief, June 6, 2007, available at http://
epic.org/privacy/ftc/google/supp_060607.pdf (``The combination of
Google (the world's largest Internet search engine) with DoubleClick
(the world's largest Internet advertising technology firm) would allow
the combined company to become the gatekeeper for Internet content. . .
. The detailed profiling of Internet users raises profound issues that
concern the right of privacy. . . .''); see also, Jaikumar Vijayan,
Google Asked to Add Home Page Link to Privacy Policies, Computerworld,
June 3, 2008, available at http://www.computerworld.com/action/
article.do?command=viewArticleBasic&articleId=9092838; Privacy
International, A Race to the Bottom: Privacy Ranking of Internet
Service Companies, Sept. 6, 2007, available at http://
www.privacyinternational.org/article.shtml?cmd%5B347%5D=x-347-553961
(We ``witnessed an attitude to privacy within Google that at its most
blatant is hostile, and at its most benign is ambivalent.'').
---------------------------------------------------------------------------
There also is a critical relationship between competition and
privacy that must not be overlooked in this discussion. Competition
ensures companies have an incentive to compete on the basis of the
privacy protections they offer. On the other hand, a dominant player
who is insulated from competitive pressure has little reason to heed
consumer demand for stronger privacy protections and faces no
significant competitive pressure from other firms offering superior
privacy practices. Indeed, if a dominant player could generate
additional profits by diluting its privacy practices, there is a
significant risk it may do so. This could bring about a ``race to the
bottom'' on privacy as other companies weaken their privacy practices
in an effort to catch up to the market leader.
Yahoo! and Google's recently announced agreement raises important
questions in this regard. Under the agreement, Yahoo! will outsource to
Google the delivery of ads appearing alongside Yahoo!'s search engine
results.\15\ This has the potential to give Google, the market leader,
further control over the sites and services where ads are served,
enabling Google to collect even more data about computer users and
potentially to combine that data with the personal information it has
on those users.\16\ It also will reduce competition in the search
advertising market, and thereby weaken Google's incentives to compete
on the quality of its privacy practices. Both of these outcomes have
implications for consumer privacy.\17\
---------------------------------------------------------------------------
\15\ See http://www.google.com/intl/en/press/pressrel/
20080612_yahoo.html.
\16\ With Google's 71 percent search query share in the U.S. based
on its relationship with AOL and Ask.com (see supra fn. 11), in
combination with Yahoo's 20.6 percent share of the core search query
market, Google will be able to gather information on up to 92 percent
of online searches. See comScore, comScore Releases May 2008 U.S.
Search Engine Rankings, June 19, 2008, available at http://
www.comscore.com/press/release.asp?press=2275.
\17\ See Jeff Chester, A Yahoo! & Google Deal Is Anti-Competitive,
Raises Privacy Concerns, May 22, 2008, available at http://
www.democraticmedia.org/jcblog/?p=596.
---------------------------------------------------------------------------
III. Microsoft's Commitment to Privacy in Online Advertising
Microsoft recognizes the role that data plays in online advertising
and the corresponding importance of protecting consumer privacy. To
guide our approach to data collection for online advertising, we
released Microsoft's Privacy Principles for Live Search and Online Ad
Targeting last July.\18\ We are deeply committed to these principles,
which focus on bringing the benefits of transparency, control and
security to the protection of consumers' data and privacy online.
---------------------------------------------------------------------------
\18\ See Appendix 1. Microsoft's Privacy Principles for Live Search
and Online Ad Targeting are also available at http://www.microsoft.com/
privacy.
---------------------------------------------------------------------------
A. Transparency
I want to first touch upon the importance of transparency.
Transparency is significant because it provides consumers with an
informed understanding of a company's data collection practices, of how
their data might be used, and the privacy controls available to users.
Without transparency, consumers are unable to evaluate a company's
services, to compare the privacy practices of different entities to
determine which online products and services they should use, or to
exercise the privacy controls that may be available to them.
Transparency also helps ensure that when consumers are dealing with a
company that has adopted responsible privacy practices, they do not
needlessly worry about unfounded privacy concerns, which could prevent
them from taking advantage of new technologies.
Transparency is also essential to ensure accountability.
Regulators, advocates, journalists and others have an important role in
helping to ensure that appropriate privacy practices are being
followed. But they can only examine, evaluate and compare practices
across the industry if companies are transparent about the data they
collect and how they use and protect it.
Transparency is especially important with respect to online
advertising. This is because consumers may not understand the types of
information that entities collect or log in providing advertisements
online. For example, many consumers may not realize that information
about the pages they are viewing, the searches they are conducting, or
the services they are using may be collected and used to deliver online
ads.
For this reason, Microsoft believes that any entity that collects
or logs any information about an individual or computer for the purpose
of delivering advertisements online should provide clear notice about
its advertising practices. This means posting a conspicuous link on the
home page of its website to a privacy statement that sets forth its
data collection and use practices related to online advertising.
Consumers should not be required to search for a privacy notice; it
should be readily available when they visit a website. This obligation
should apply to entities that act as ad networks, as well as to
websites on which ads appear--whether they display ads on their own or
rely on third parties to deliver online advertising.
In addition to being easy to find, the privacy notice must be easy
to understand. While many websites have publicly posted a privacy
notice, this alone is not enough. Too often, the posted privacy notice
is complex, ambiguous and/or full of legalese. These notices make
privacy practices more opaque, not more transparent. Instead, short and
simple highlights are essential if consumers are to easily understand a
company's information practices. It helps avoid the problem of
information overload, while enabling consumer awareness.
Finally, to ensure that the consumer can be fully informed, the
privacy notice should also describe the website's data collection and
use activities in detail. This includes, at a minimum, descriptions of
the types of information collected for online advertising; whether this
information will be combined with other information collected from or
about consumers; and the ways in which such information may be used,
including whether any non-aggregate information may be shared with a
third party.
Microsoft has embraced these obligations. We post a link to our
privacy notice on every page of our websites, including the home page.
We also were one of the first companies to develop so-called
``layered'' privacy notices that give clear and concise bullet-point
summaries of our practices in a short notice, with links to the full
privacy statement for consumers and others who are interested in more
detailed information. And our privacy statement is clear about the data
we collect and use for online advertising. Further, we have released
more detailed information about our practices, such as a white paper
that describes the methods we use to ``de-identify'' data used for ad
targeting.\19\ To illustrate our efforts to be transparent about our
practices, we have included in Appendix 2* screen shots of
the privacy link available on the home page of our Windows Live search
service and of our layered privacy notice, including both the short
notice and our full online privacy statement.
---------------------------------------------------------------------------
\19\ See section III.C below.
---------------------------------------------------------------------------
B. Control
The second core principle Microsoft looks to in protecting our
customers' privacy is user control. Consumers should have a choice
about how information about their online activities is used, especially
when that information can be aggregated across multiple websites or
combined with personal information. Microsoft has made consumer control
a key component of our practices online.
As an example, Microsoft has recently deployed a robust method to
enable users to opt out of behavioral ad targeting. As background, most
industry players that offer consumers a choice about having information
about their online activities used to serve behaviorally targeted ads
do so by offering consumers the ability to place an ``opt-out'' cookie
on their machines. In general, this process works well, but it does
have some inherent limitations. For example, opt-out cookies are
computer-specific--if a consumer switches computers, he or she will
need to specify any opt-out preferences again. Further, if cookies are
deleted from the user's PC, that user's opt-out choice is no longer in
effect. To address these limitations, Microsoft now gives consumers the
option to tie their opt-out choice to their Windows Live ID. This means
that even if they delete cookies on their machine, when they sign back
in their opt-out selection will persist. It also means that a single
choice can apply across multiple computers that they use. This will
help ensure that consumers' choices are respected.\20\
---------------------------------------------------------------------------
\20\ Microsoft's personalized advertising opt-out page is available
at https://choice.live.com/advertisementchoice/Default.aspx.
---------------------------------------------------------------------------
Microsoft also has committed to respecting consumers' opt-out
choice on all sites where it engages in behavioral advertising. This
means that consumers are offered a choice about receiving behaviorally
targeted ads across both third-party websites on which Microsoft
delivers behaviorally targeted ads, as well as Microsoft's own
websites. This is important because consumers reasonably expect that
the opt-out choice offered by a company would apply on all websites
where that company engages in behavioral advertising practices. This is
another example of where we have committed to going beyond standard
industry practice to better protect the interests of consumers.
We also recognize it is appropriate that the level of consumer
control may vary depending on the data that will be used to serve an
online ad. For example, many consumers have serious reservations about
the receipt of targeted advertising based on the use of certain
categories of personally identifiable information, particularly those
that may be considered especially sensitive. Thus, we have proposed
that companies should obtain additional levels of consent for the use
of such information for behavioral advertising--including affirmative
opt-in consent for the use of sensitive personally identifiable
information.\21\
---------------------------------------------------------------------------
\21\ See, for example, Microsoft's comments to the Federal Trade
Commission's proposed self-regulatory framework for online advertising,
included as Appendix 4* and available at http://www.ftc.gov/
os/comments/behavioraladprinciples/080411microsoft.pdf.
---------------------------------------------------------------------------
C. Security
The third principle we look to in protecting consumers' privacy is
that strong, simple, and effective security is needed to strengthen
consumers' trust in our products, the Internet, and all information
technologies. Security has been fundamental at Microsoft for many years
as part of our Trustworthy Computing initiative. And it plays a key
role with respect to our online advertising practices.
We have taken a broad approach to protecting the security of
computer users with respect to serving ads online. This approach
includes implementing technological and procedural protections to help
guard the information we maintain. We also have taken steps to educate
consumers about ways to protect themselves while online, and we have
worked closely with industry members and law enforcement around the
world to identify security threats, share best practices, and improve
our coordinated response to security issues.
In addition, we have designed our systems and processes in ways
that minimize their privacy impact from the outset while simultaneously
promoting security. For example, we use a technical method (known as a
one-way cryptographic hash) to separate search terms from account
holders' personal information, such as name, e-mail address, and phone
number, and to keep them separated in a way that prevents them from
being easily recombined. We have also relied on this method to ensure
that we use only data that does not personally identify individual
consumers to serve ads online. As a result of this ``de-
identification'' process, search query data and data about Web surfing
behavior used for ad targeting is associated with an anonymized
identifier rather than an account identifier that could be used to
personally and directly identify a consumer.\22\
---------------------------------------------------------------------------
\22\ A white paper describing Microsoft's ``de-identification''
process is attached to these comments as Appendix 3.* It is
also available at http://www.microsoft.com/privacy.
---------------------------------------------------------------------------
Finally, we have implemented strict retention policies with respect
to search query data. Our policy is to anonymize all such data after 18
months, which we believe is an appropriate time-frame in our
circumstances to enable us to maintain and improve the security,
integrity and quality of our services. We intend to continue to look
for ways to reduce this time-frame while addressing security, integrity
and quality concerns. In addition, unlike other companies, our
anonymization method involves irreversibly removing the entire IP
address and other cross-session identifiers, such as cookies and other
machine identifiers, from search terms. Some companies remove only the
last few digits of a consumer's IP address, which means that an
individual search query may still be narrowed down to a small number of
computers on a network. We think that such partial methods do not fully
protect consumer privacy, so we have chosen an approach that renders
search terms truly and irreversibly anonymous.
IV. Microsoft's Support for Self-regulation and Privacy Legislation
Microsoft believes that these core principles of transparency,
control, and security are critical to protecting consumers' privacy
interests online. These principles form the basis for our support of
robust self-regulation in the online advertising market and for
baseline privacy legislation.
We have been an active participant in self-regulatory efforts.
Microsoft has been engaging with the Network Advertising Initiative
(``NAI''), a cooperative of online marketing and advertising companies
that addresses important privacy and consumer protection issues in
emerging media.\23\ The NAI is currently in the process of revising its
guidelines to address changes in the online advertising industry. The
NAI's efforts have been critical to understanding the privacy issues
associated with online advertising, and we will continue to work with
them as they finalize their draft proposal.
---------------------------------------------------------------------------
\23\ Atlas, which was part of Microsoft's recent acquisition of
aQuantive, was a founding member of NAI.
---------------------------------------------------------------------------
We also filed comments responding to the Federal Trade Commission's
request for input on a proposed self-regulatory framework for online
advertising. In our comments, we explained the need for a broad self-
regulatory approach since all online advertising activities have
potential privacy implications and some may be contrary to consumers'
expectations. To this end, we proposed a tiered approach to self
regulation that is appropriately tailored to account for the types of
information being collected and how that information will be used. It
would set a baseline set of privacy protections applicable to all
online advertising activity and would establish additional obligations
for those companies that engage in practices that raise additional
privacy concerns. We are attaching a copy of our comments to the FTC
for your convenience.\24\
---------------------------------------------------------------------------
\24\ See Appendix 4. Our comments are also available at http://
www.ftc.gov/os/comments/behavioraladprinciples/080411 microsoft.pdf.
---------------------------------------------------------------------------
In addition to supporting self-regulatory efforts, we have long
advocated for legislation as a component of effective privacy
protections. We were one of the first companies to actively call for
comprehensive Federal privacy legislation.\25\ More recently, we have
supported balanced and well-crafted state legislation on privacy in
online advertising that would follow the general structure proposed in
our FTC comments.\26\ And we would be glad to work with the Committee
on similar national privacy standards that would protect both privacy
and opportunities for innovation in the online advertising industry.
---------------------------------------------------------------------------
\25\ See http://www.microsoft.com/presspass/download/features/2005/
PrivacyLegislationCall
WP.doc.
\26\ A. 9275-C, 2007-2008 Reg. Sess. (N.Y. 2008), available at
http://assembly.state.ny.us/leg/?bn=A09275&sh=t (imposing minimum
notice and choice obligations on certain website publishers and
advertising networks); S. 6441-B, 2007-2008 Reg. Sess. (N.Y. 2008),
available at http://assembly.state.ny.us/leg/?bn=S06441&sh=t (imposing
baseline notice, choice, security, and consumer access obligations on
certain third-party advertising networks); H.B. 5765, 2008 Gen. Assem.,
Feb. Sess. (Conn. 2008), available at http://www.cga.ct.gov/2008/FC/
2008HB-05765-R000148-FC.htm (imposing minimum notice, choice, security,
and use limitations on third-party advertising networks).
---------------------------------------------------------------------------
Our support of self regulation in the online advertising market and
prudent privacy legislation is only a part of our comprehensive
approach to protecting consumer privacy. We will continue to support
consumer education efforts to inform users of how to best protect
themselves and their information online. And we will persist in our
efforts to develop technology tools that promote the principles of
transparency, control, and security. In short, we are prepared to work
collaboratively on all fronts to maintain the growth of online
advertising while fostering consumer trust online.
V. Conclusion
Microsoft recognizes that the protection of consumer privacy is a
continuous journey, not a single destination. We can and will continue
to develop and implement new privacy practices and protections to bring
the benefits of transparency, choice, and security to consumers. Thank
you for giving us the opportunity to testify today. We look forward to
working with you to ensure consumers' privacy interests are protected
as they continue to enjoy the proliferation of free services and
information that online advertising supports.
* The information referred to has been retained in
Committee files.
Senator Dorgan. Mr. Hintze, thank you very much. I
appreciate your testimony.
To my colleagues, I would say time is not our friend this
morning. The vote is supposed to start at 11:15, although I am
told it may slip. We will probably know shortly. If it starts
at 11:30, that means that we would have perhaps until 11:45
before we would have to depart this room.
I will do 5-minute rounds here, and if we finish at 11:45,
we will not have to come back. We have five votes in
succession, which means we probably would not be able to come
back until 1 o'clock. So my hope would be that for the next 45
minutes to an hour, we will all be able to have an opportunity
to ask relevant questions.
And I thank very much the witnesses for being here.
I have 100 questions, regrettably. Let me take just 4 and a
half or 5 minutes and then call on my colleagues.
First of all, online advertising is helpful and useful in
my judgment. I understand that. It helps support the Internet
itself, which has some wonderful companies and sites providing
useful information services, entertainment. I understand all of
that.
The question today is not, as Mr. Crews indicated, are
certain kinds of advertising the devil. I think advertising is
a necessary component of the Internet and is helpful to
consumers. The question is about the collection of information
about consumers as they travel the Internet.
And Mr. Dykes, I will ask you the first question. The
stories that I have seen about maybe an Internet service
provider deciding we are going to allow NebuAd to come in, and
whenever anybody does anything on our system, as someone who
has signed up for our Internet service provider service, we are
going to essentially shovel all that information over to you as
it is being done. I mean, what is the difference between that
and tapping into somebody's wire, using the pejorative term
``wiretapping''? Is that not just wiretapping?
Mr. Dykes. No, sir. We believe that we are not violating
the wiretap law. I am not a lawyer, but my lawyers have told me
we are in compliance with the law, and they have prepared a
memo on the subject and I would be prepared to submit that for
the record. [The information requested is published on pp. 101-
107 of this transcript.]
But it is important to note that the information that we
are looking at as people surf the web does not involve any
personally identifiable information. All we are doing is taking
an anonymous identifier. We are taking their IP address, for
example, and transforming that into an anonymous number with a
one-way hash. And against that anonymous identifier all we are
examining is qualification for market segments. So we are not
keeping the raw data. It is just qualification for market
segments against an anonymous identifier.
Senator Dorgan. But your approach and the approach of the
Internet service provider would not be an opt-in approach. It
would be an opt-out approach. I would think if my Internet
service provider said to me, you know what, Mr. Dorgan? We have
a proposition. Is it OK if we give everything you do to another
company? I would say, of course, it is not OK. You kidding me?
The answer is no, capital N-O. So from an opt-in standpoint, I
am guessing that this would not be a workable model. It only
works if you require people to opt-out. I mean, I think that is
the difficulty.
Ms. Harris, do you want to comment on that?
Ms. Harris. Yes, I do. I think, first of all, it is
important to understand that our wiretap laws do not care if
the information is personally identifiable or not personally
identifiable. The laws are agnostic on that point. It is
important to understand that.
Second, they may not be using all the information, but they
are mirroring. They are capturing the data stream and then
somehow mirroring or copying it. So the ISPs are providing that
information to an unknown third party, a man in the middle. I
take Mr. Dykes' word that they are then not using all of that
information, but you have to have a way to separate out the
information you are using from the information you are not
using. I do not think you can pretend----
Senator Dorgan. Mr. Dykes, do you wish to respond to that?
Mr. Dykes. Yes, I would, Mr. Chairman. So although the
information flows through our system, information that does not
conform to one of the market segments we are looking at is
simply ignored and flushed permanently. And so we are only
looking at these select innocuous market segments.
I would also point out that we do provide very robust
notice to the user and an opportunity to opt out. These notices
are sufficiently robust that people do opt out. So it is not
like people are ignoring them. They are informed, and as I
said, we are confident that we do not break the law and we have
a memo on the subject that we would submit.
Senator Dorgan. Mr. Kelly, on Facebook, I do not know
whether this has changed, but my understanding is that when
someone would order an application, which they could on
Facebook, called Dogbook or Scrabble, that that application
then would allow the person selling the application access to
all that which exists in Facebook. Was that the case?
Mr. Kelly. So, first of all, a user has to affirmatively
add an application to a profile and----
Senator Dorgan. I understand, but when they do that----
Mr. Kelly.--box that informs them that they are adding this
application and sharing information with the third party in
that case. At that point, the application has the ability to
request certain data. We do not hand over all the data at that
point. They have the ability to request it in a session basis
and limit it to the data that the user who has installed the
application can see on Facebook. So the privacy settings
persist.
That application is allowed to, for caching purposes,
retain the information they have requested for only 24 hours,
and if they exceed that amount, we have a policy enforcement
team which will take action, including potentially shutting
down the application. And in the last few weeks, we have shut
down a number of applications over violations until they come
into compliance.
Senator Dorgan. I have just received some good news. The
U.S. Senate was not constructed to ever be accused of speeding,
and all of us who serve here know that. It actually has
exhibited that this morning. We have actually an 11:45 start
time, which means we have a full hour for questions. The 11:45
start time will be the first vote. So that is good news.
Let me call on Senator Vitter.
STATEMENT OF HON. DAVID VITTER,
U.S. SENATOR FROM LOUISIANA
Senator Vitter. Thank you, Mr. Chairman. I have some
general questions and I would invite very concise responses
from anyone who has a response.
What is the best estimate available regarding the use of
this type of information not for behavioral advertising--put
that on the side--but for other purposes that all of us would
consider an abuse, identity theft or other abusive uses, not
behavioral advertising? Does anyone have a sense of the size of
that problem?
Ms. Harris. Mr. Vitter, I think that the answer is nobody
knows and there are no rules in place that would prevent any of
that. And so the question is, and I will say that for members
of the Network Advertising Initiative that they have made a
commitment not to do that, but it is a big Internet. And we do
not know what everybody is doing. But basic fair information
practices say that if you collect information for a purpose,
you should use it for that purpose, and not use it for another
purpose. We have no legal way of enforcing that.
Senator Vitter. Does anyone else have a sense of the size
of that problem?
Mr. Dykes. Well, I would point out that when NebuAd was
being founded back in 2006, it was when AOL search data became
public and people determined that a large amount of raw search
data could represent personally identifiable information. We
also had the Government asking Verizon and AT&T to provide
click stream data.
And it was for those reasons that when NebuAd was founded,
we resolved never to be in a position to have such data that,
which if found by others, could ever have a problem. That is
why we do not keep the raw data mapped against a user ID. We
only keep this qualification for market segments and all our
user ID's are anonymized. So we keep very limited data to avoid
any of those risks.
Senator Vitter. I understand.
Does anyone else have a sense of the global size of that
problem? OK.
Now, with regard to that problem--again, put behavioral
advertising on the side. With regard to that problem, would
ensuring that all of this collection of data is made anonymous
in the ways that several of you do presumably already solve
that problem or not?
Ms. Harris. I do not think you can do that. The industry
wants us to believe that this information is anonymous. I think
at best it is pseudo-anonymous. They are building profiles. And
I think a couple of years ago when AOL released search data for
a good purpose, for research purposes, it took very little time
for people to take a single set of search data and identify
somebody. So that we are moving to a point where there is
enough information----
Senator Vitter. Can I interrupt for a second?
Ms. Harris. Yes.
Senator Vitter. Explain that to me as a layperson because
my reaction to that is there are a gazillion people on the
Internet. How do you possibly take that anonymous information
and come up with an individual?
Ms. Harris. Well, it cannot be entirely anonymous because
you are trying to put--I mean, even for NebuAd--and we can talk
about this--they keep refreshing the information about an
individual. You have to. Right?
Mr. Dykes. We only keep information in market segments. The
way AOL's data became identifiable is that people did numerous
searches on houses in their neighborhood, and soon it was
fairly clear who the person was who was doing the search
because there was just so much data centered on a particular
name and address and things like that.
Ms. Harris. But profiles can include that. Profiles can
include that you have been on a particular website looking at a
particular thing, that you have searched for your own name,
that you--it depends what is in the profile.
Mr. Dykes. Exactly. If the profiles had that information,
that is when it becomes pseudo-anonymous, and therefore, you
can derive PII.
Senator Vitter. Let me reask the question. Could there be a
regime ensuring true anonymity in terms of the collection,
number one? Is that possible?
Mr. Dykes. I believe so.
Senator Vitter. And number two, is any legitimate purpose
that is of any arguable benefit to consumers sacrificed through
that regime?
Mr. Dykes. Well, the effectiveness does decline as you move
toward eliminating pseudo-anonymous data. But we have chosen to
make that choice. We do not have very sensitive ads. We have
chosen not to take very sensitive ads from big pharmaceutical
companies, for example. There are a lot of choices we have made
to protect consumer choices that do reduce the economic value,
but these I think are important choices.
Senator Vitter. Does anyone else have a reaction to that
idea of what most folks would regard as true anonymity? Number
one, how possible it is; number two, what if anything would be
sacrificed.
Mr. Crews. I think you are always taking a risk when you
think that the Internet, with the kind of network that it is--
if pure privacy is what you want, the Internet is probably the
wrong network for you because everybody here has a business
card, and the people that we talk about with most of these new
technologies--and I talked about warring computer scientists.
Some are trying to offer anonymity.
But on the other hand, there are going to be cases where
you do not want somebody to pose as you and you are going to
want to use personally identifiable information. The
technologies are going to change as biometrics get integrated
into commerce and things like that. So we are not going to want
anonymity ultimately in some respects, if we are dealing with
our insurance company, dealing with a finance company.
Senator Vitter. Let me back up. I am not talking about
anonymity dealing with your insurance company. I am talking
about anonymity in terms of the collection of data for the
purposes that we are talking about.
Mr. Crews. I do not think so because of the openness of the
Internet and the criminal element that is always going to be
there. You know, anybody can go into Starbucks and the library
and get on the Net. It is always going to be an open network
that nobody has a proprietary stake and can offer that kind of
a guarantee. We are going to have to have activities on the
consumer side and institutions like identity theft insurance
and all those sorts of things too. We can try our best, and I
think that is what scientists are trying to do. But the
Internet is not the network for privacy in a sense.
Senator Vitter. Thank you.
Senator Dorgan. Senator Stevens was here and he had to go
to an Appropriations Committee markup, and he asked that we put
his statement in the record. We will do that by unanimous
consent.
[The prepared statement of Senator Stevens follows:]
Prepared Statement of Hon. Ted Stevens, U.S. Senator from Alaska
Mr. Chairman, thank you for scheduling this hearing.
As we recently learned in last month's hearing on spyware, we must
be increasingly vigilant in ensuring that Americans' personally
identifiable information is protected. On one hand, legitimate online
advertising provides many benefits to our economy--for example in
Alaska it helps our newspapers, radio stations and television stations
to provide free online news. But, on the other hand there are concerns
about protecting individuals' privacy and preventing against identity
theft.
For the Internet economy to continue to grow, Americans need to
have confidence that their personal identifiable information is safe
when they enter their data online. Moreover, consumers should be fully
informed about what information is being collected online, who is
collecting the data, and what options consumers have to protect
themselves.
At this hearing, it will be important that the Committee gets a
sense of how online behavioral advertising works, and how consumers can
guard their personally identifiable information. I also anticipate that
today's hearing will help us better understand what roles the Federal
Trade Commission and the industry should play in educating consumers on
how information is collected online and how to protect themselves.
I thank the witnesses for participating today and look forward to
hearing their testimony.
Senator Dorgan. Senator Klobuchar?
STATEMENT OF HON. AMY KLOBUCHAR,
U.S. SENATOR FROM MINNESOTA
Senator Klobuchar. Thank you very much, Senator Dorgan, for
holding this hearing, and thank you to our witnesses. I was
thinking, as I listened, how everyone sort of has a love-hate
relationship with advertising on the Internet. I love it when
it is something for a discount on clothes, and I hate it when
my daughter sees the American Girl things put on the screen.
But I think everyone feels like that.
And I think we also know that advertising plays an
important role in enhancing a consumer's online experience. It
is revenue, and it also promotes the development of online
news. So like Senator Dorgan, I would agree that we are not
against advertising on the Internet, but the issue is, as it
becomes more sophisticated, do we have a role here to play in
making sure that consumers' privacy is protected as companies
develop more technology and are able to dig deeper into that
information, that we keep it anonymous.
I guess my first questions are of you, Ms. Parnes. I know
Mr. Kelly and Mr. Hintze were talking about how the FTC could
go farther in terms of differentiating between the kind of
information, whether it is anonymous information, whether it
is--I think they were differentiating between the way you get
the information based on personally identifiable information or
advertising based on non-personally identifiable information.
Could you do that as part of your rules?
Ms. Parnes. Well, Senator, I think we could make those
distinctions. In the first instance, let me just clarify that
what the Commission put out was a set of proposed principles
that would guide----
Senator Klobuchar. Self-regulation.
Ms. Parnes.--a self-regulatory scheme.
Senator Klobuchar. Right. They were not rules.
Ms. Parnes. Not Commission rules.
But within the the principles, I think our proposal
reflects some differences. For example, the proposal calls for
express affirmative consent before an entity can collect
sensitive personal information from a consumer for use in
behavioral advertising. So we acknowledge that there are some
areas that are more sensitive and that require a certain level
of heightened transparency and control and choice for
consumers. I think that as we are looking at all the comments
that we received, that is certainly something that we will
consider, whether we should provide a sliding scale type of
scheme.
Senator Klobuchar. The other thing is I know you have had
60 comments, and I am sure a lot of them have been industry
groups and privacy groups. But again, as Senator Dorgan was
mentioning, I am not sure individuals out there even understand
what is going on.
I was reading about the European Union has some guidelines
where you have to be able to access what people have on you and
then correct it if it is false. Have you looked at that at all?
Ms. Parnes. In these sets of principles, we have not looked
at access to the information. I think what we have been more
concerned about is the issue of transparency, consumers really
understanding what is going on in behavioral advertising.
I frankly was surprised that among the over 60 comments we
received, we did get some comments from individual consumers,
but I think that you are absolutely right. For the most part,
consumers do not understand what is happening behind the
screen, as it were.
Senator Klobuchar. Mr. Hintze, you talked about the need
for a Federal law, and I think Ms. Harris talked about updating
Federal laws. What ideas do you have for changing Federal law
that we could look at?
Mr. Hintze. Well, going back a couple of years to 2005, we
came out in support of a comprehensive Federal privacy
legislation. We think that establishing a baseline at the
Federal level across industries makes a lot of sense. It would
require certain basic requirements that are consistent with
good industry practices and self-regulation that exists today
around user notice and security and user consent and even
access, as you have mentioned. There are some very difficult
issues there. We have some ideas of what a good Federal law
might look like.
We have also been supportive of State legislation in some
cases where it was balanced and provided appropriate guidelines
for the online advertising space. We have also been very
supportive of the security breach notification laws and others.
But I think that overall, legislation does play an
appropriate part of the mix of protecting consumer privacy
along with self-regulation and consumer education and
technology tools.
Senator Klobuchar. So the idea would be to take some of the
things that Ms. Horvath has been talking about, some of these
other things and actually put them into law. How would it work
with the self-regulation?
Mr. Hintze. Well, the Federal law would presumably provide
a baseline common set of requirements, and then for particular
industries, for example, there may be additional very specific
self-regulatory additions that could go on top of that. I think
online advertising is a perfect example of that where there are
some specific aspects of online advertising around behavioral
targeting and what consent looks like in those cases. That may
be more appropriate for self-regulation, but could certainly
build on top of the baseline requirement that is based in law.
Senator Klobuchar. Mr. Kelly, did you want to add, since
you raised that issue of different ways that we could
differentiate between information? Are you in favor of some
kind of Federal law?
Mr. Kelly. We have not taken a position on a formal Federal
law around that. We have been focused on creating innovative
privacy technology to allow users to control the collection and
use of data. An instructive example here would be that, for
instance, in a classic behavioral targeting situation, an ad
network would see that you visited five auto sites and
determine from that that you might be interested in buying a
car. Whereas, the Facebook approach has been, let us say, if we
were, for instance, partnered with an auto site, the auto site
would allow you to say--you know, give you a little pop-up that
said, do you want to share the fact that you purchased a Chevy
with your Facebook friends, and give you a real-time choice
around that. So we are focusing on technology as a solution.
Senator Klobuchar. Thank you very much.
Senator Dorgan. Senator Klobuchar, thank you very much.
Senator DeMint?
STATEMENT OF HON. JIM DeMINT,
U.S. SENATOR FROM SOUTH CAROLINA
Senator DeMint. Thank you, Mr. Chairman. Thank you for
holding the hearing. I appreciate your comments on the
importance of advertising. I think we all agree that the
ability to get all the free things we get on the Internet are
important, and they do relate directly to advertising. And I
think probably most of us would agree that the ability of
advertisers to target their markets are important to continue
to attract advertising revenue to the Internet, and I think we
have heard a lot about how the industry has progressed.
Just a couple of questions here. Mr. Dykes, just as a for
instance so I can kind of get it, this anonymous versus not
anonymous question, if the Department of Justice were to issue
you a subpoena asking you for names of people who have searched
for information about explosives, would you be able to provide
the names or locations of individuals who had done that?
Mr. Dykes. No, sir, we would not be able to provide names,
nor even their IP addresses. We have no personally identifiable
information in our system, and it would just simply not be
possible to get back to an individual from any information we
have in our system.
Furthermore, we do not have advertisements and categories
on bomb-making, for example, and therefore, that information is
not even in our system either.
Senator DeMint. So it appears that the technology has been
developed and is being developed and improved that would allow
increasing privacy and anonymity on the Internet, which is I
think impressive, given the fact that this is all being done
voluntarily.
It would be my assumption that the incentives for--the
businesses represented at the table have a lot of incentives to
compete for the best privacy policies, that they would disclose
their privacy policies to encourage users to use their
particular service because of their ability to protect. I think
the private market has a lot of incentives.
Let me just direct a few questions at Ms. Parnes. How long
do you think the rulemaking or even the development of these
principles might take?
Ms. Parnes. Well, I am reluctant to give a specific time-
frame, but we are moving quickly on this. As we noted, we
received a lot of comments, and this is a challenging issue. We
want to drill down, figure out what information we have, if we
need additional information, and then move forward, as I said,
very quickly.
Senator DeMint. Do you think at least a year or 2?
Ms. Parnes. I would doubt that.
Senator DeMint. Do you think it is the responsibility of
the FTC to regulate advertising on the Internet?
Ms. Parnes. Well, in some ways we do regulate advertising
on the Internet. The question about behavioral advertising--I
think some of our principles are grounded in section 5, and we
do have principles that govern behavioral advertising, as well
as all other advertising.
Senator DeMint. It would be my hope that you continue to
see your responsibility as protecting consumers and not
necessarily attempting to manage or regulate different aspects
of our economy.
I think in some ways we have got a solution in search of a
problem, as the industry moves very quickly to try to cut these
problems off before they occur. It is very likely that if
rulemaking or even principle-making is going to take a year or
2--and obviously, the FTC needs to look at it--that the
technology that is being talked about today will advance--will
make great strides over the next months and years. And it is
very likely that by the time the FTC acts, that the industry
will be far ahead of where you are.
My concern is this. The Government's attempt at privacy
certainly is well motivated, but rules, for instance, in the
health care industry where, when I was with a family member
having surgery, the physician was not able to give me
information about actually what took place or I could not call
a doctor's office and get information about my own daughter's
bill. And you hear doctors talk about the inefficiencies that
are created because of this well intended policy. I just have
great concerns that if the Federal Government attempts to jump
in in the middle of a situation where a problem has not yet
exhibited itself, that we are likely to inhibit one of the
showcases of free enterprise in America today.
Ms. Parnes. Senator, I would just say that that is
precisely why the Commission is encouraging self-regulation in
this area. The principles that we have proposed are principles
that would govern a self-regulatory scheme that industry itself
adopts, and we think that self-regulation, at least at this
point, is more appropriate in this kind of environment where,
as you note, technology is changing so quickly.
Senator DeMint. Well, that is a good philosophy, and I
appreciate hearing it this morning. And I thank all the
panelists.
Mr. Chairman, I yield back.
Senator Dorgan. Senator DeMint, thank you very much.
Senator Thune?
STATEMENT OF HON. JOHN THUNE,
U.S. SENATOR FROM SOUTH DAKOTA
Senator Thune. Thank you, Mr. Chairman. I too want to thank
you for holding the hearing. I think this is an issue that is
very important. It is being talked about, getting a lot of
attention. It is going to get more attention, I think, in the
future. I do believe that it is critical that consumers have
knowledge of whom or what is tracking their online activity and
additionally that they have every opportunity to learn more
about the uses of their personal information.
In that light, I have a question for Ms. Parnes, and that
has to do with sensitive information which certainly, I
believe, would deserve a greater degree of protection than just
regular online uses. What do you consider to be sensitive
personal information? Is that a health record, a credit card
history, those sorts of things?
Ms. Parnes. In our proposal, we identified several
examples: medical information, financial information, and any
information about children. It is not intended to be an
exclusive list, though.
Senator Thune. What is the relationship between market
share in the search advertising business and the ability to
attract advertisers and web publishers? And is there a tipping
point where a company, for example, say a Google, could become
so dominant in the market that advertisers and web publishers
have no choice but to contract with the company?
Ms. Parnes. I am afraid that----
Senator Thune. Of venturing into that?
Ms. Parnes.--I do not think I really have the information.
Yes. It would be difficult to venture a guess on that. Perhaps
one of the people here might want to.
Senator Thune. All right. We will let you off the hook, but
if anybody else would care to take a stab at that.
Mr. Crews. Well, I do not think you have to worry so much
about tipping points in the online world for a lot of reasons,
but in particular, as long as people have the ability to click
that button and there is no censorship online--you know,
Government does not dictate where we go or anything like that--
no one can attain that kind of monopoly power because nobody
operates in a vacuum.
The joke way I put it when debating media ownership is that
even if the biggest companies merge with others across other
avenues, what happens? They could get together and try to abuse
the trust or what have you. But what happens? Advertisers flee.
Wall Street flees. Customers flee. The employees flee. The
people who work in that business who know the technology and
the science move and start another company.
So all of the impulses in a market economy, especially in
one where barriers to entry are so low, like in Internet
businesses, there is not that kind of threat of monopoly power.
It was not that long ago that all of us were using Yahoo! to
search or Alta Vista to search, and then Google came along with
another technology. It is one algorithm, but you can go on the
search engine Colossus I believe. There are hundreds of search
engines out there. Just a thought.
Senator Thune. If Congress were to establish regulatory
guidelines or core principles for online advertising privacy,
what would those principles entail, and how would those align
with some of the principles that the FTC is adopting? Does
anybody want to venture----
Ms. Harris. Yes. Mr. Thune, I think it is important to note
CDT's position in support of a baseline privacy law is not
about advertising. It is a step down. It is for all data
collection and use practices. So those of us who are advocating
adoption, I think you are looking at fair information
practices. You collect for a particular use. You hold it only
for the time that use exists. You do not hold data forever.
People have a right to know what people are holding about them.
They have a right to opt out or opt in, depending on the
sensitivity of the data. The the devil here is in the details,
but I do not think that the framework--you know, fair
information practices is sort of infused throughout our legal
system and certainly in terms of Government information,
although somewhat outdated.
Let me just add one thing. We do understand and have no
interest in slowing down the technological innovation in terms
of privacy-enhancing technologies. We do think that that is an
important part of this, and we would not want to see any law
that froze the Internet in a way that would make that
impossible. So everything is a balance here.
Mr. Hintze. I think that is right. We generally agree, and
one of the reasons that we came out in favor of comprehensive
Federal privacy legislation in 2005, which was before we were
significantly involved in the online advertising space at all,
is that there is just this emerging patchwork of Federal and
State privacy laws with different standards depending on the
industry, depending on the particular issue. And it just made a
lot of sense to harmonize that, both from a compliance
standpoint from a company's point of view and from a consumer
standpoint so they can know that they have a common baseline
protection across the board, regardless of the kind of industry
that they are dealing with.
As we get into online advertising, that kind of baseline
Federal privacy standard could certainly help here. It does not
answer all the questions, and that is why we need to supplement
it with self-regulation and other practices. But I think the
need for a Federal law is not specific to online advertising
but is more general.
And I would just like to make one other point about your
other question earlier about the competition and dominance in
this space. I think for the purposes of this hearing, we
definitely see a nexus between competition and privacy in that
if there does become a dominant player in the online
advertising space, that means that there is a single company
that is collecting more data. The more ads you serve, the more
data you are collecting. And it could become the case where a
single company has a nearly complete picture of people's online
behavior.
Beyond that, we also think that the need for competition in
the space fosters competition in privacy. If there is a
dominant player, there is little or no incentive to adopt
better privacy practices because they are not facing
competitive pressure from other companies that may be adopting
superior privacy practices.
Mr. Dykes. Senator, I would like to agree with both of the
prior speakers. I think the law should be focused on privacy
and what types of information are being collected. And they
should be technology-neutral and business process-neutral and
not necessarily advertising-focused either. It should really be
focused on privacy.
But I also think that the Government needs to be very
careful in making policy to ensure that it does not stifle
competition for the very reasons we just noted because there
are new competitors springing up all the time. When we think we
have one dominant competitor, past history has shown that there
are other ones in the wings springing to life, and we would not
want the Government to stifle that. So there is a role here for
self-regulation as well to allow for flexibility as technology
develops.
Ms. Harris. Can I just make one brief point on this notion
of technology neutrality in the law? We do have neutrality
under ECPA. If you are acting as a conduit, you are treated one
way. If you are not you are treated the other. And I think we
have to be careful. We have wiretap laws. We have other
communications laws that apply to the institutions that stand
between the user and the ends of the network. I think we can
sort of use this neutrality--everybody should be treated the
same at all times--as a mantra and lose, I think, track of the
fact that companies stand in different positions to a consumer,
and we have to take that into account.
Senator Thune. Thank you all for your testimony.
Thank you, Mr. Chairman.
Senator Dorgan. Senator Thune, thank you.
Senator Nelson?
STATEMENT OF HON. BILL NELSON,
U.S. SENATOR FROM FLORIDA
Senator Nelson. Mr. Chairman, I think the timing of this
hearing is uncanny because we have on the floor right now the
FISA bill which the whole question in FISA is about the
protection of the rights of privacy and the civil liberty of
Americans. I can tell you all that as a member of the
Intelligence Committee, we have been struggling with this FISA
issue for the last year and a half. We have come up with a
compromise and a resolution of how you can allow our
intelligence agencies to seek the information of a terrorist
but at the same time protect the civil liberties of American
persons, which is defined as an American citizen and a person
not an American citizen who is legally within the United
States.
What I am struck with is that we have a similar issue here.
Take, for example, I use the Internet to go online to read the
newspapers back home in Florida. Now, if suddenly the kinds of
articles that I am reading are going to be determined, the
content of what I am reading is going to be identified with me
so someone can target advertising, I want to question the
underlying basis of this.
In our discussion of snooping of terrorists before, we have
carved out an exception that we do not want the Government--
now, here we are talking about the private sector, but we do
not want the Government sector to go and examine what books
that we are reading at the local libraries. Well, right here we
have the question of whether or not we are going to let other
private people within the private sector examine the same thing
and then use it for a commercial advantage.
Now, I am not naive enough not to understand that this is
already happening. So the question is, how do we rein this in
so that as this Internet continues to explode into something
years down the road that we cannot even conceive now that will
be totally ubiquitous, how do we support our Constitution and
protect our civil rights?
So let me just ask, is there a way that we can approach
this where we would govern the type of Internet connection used
instead of the content or the information collected? What do
you think about that, Ms. Harris?
Ms. Harris. Well, I think I am not entirely sure of what
you are suggesting, Senator.
Senator Nelson. Nor am I.
[Laughter.]
Ms. Harris. Well, I am not sure that that is really the
answer.
I will say this, that you are, I think, very prescient to
connect what appears to be a totally unrelated matter of FISA
with what is going on here. And the reason is that when we
operate in an Internet regime where we have no limits on how
long a company can collect data, where we have these other sets
of laws, you know, the Electronic Communications Privacy Act,
et cetera, that are 20 years old and did not anticipate the
Internet. Whatever we are building here and whatever is being
collected, to the extent that it can be pseudo-anonymously
connected through user logs, et cetera, it is also available to
the Government. So I do not think we can view these things as
there is a consumer stream that is being collected out here and
there is a Government stream.
If you look at e-mail, for example, we have very little
protection in this country for stored e-mail because we did not
anticipate 20 years ago that it was going to be stored online.
That information has very little legal protection. So we are in
a perfect storm----
Senator Nelson. Yes, we are.
Ms. Harris.--between the commercial activity and national
security----
Senator Nelson. Yes, just like--you know, we hold it very
dear in this country that we do not like other private citizens
reading our personal mail, but in effect, what we have is the
ability to read our personal communications here by other
private individuals separating the issue from Big Brother and
the Government.
Ms. Harris. I do not think you can.
Senator Nelson. Well, let me ask you this, Ms. Harris.
Ms. Harris. That is not to suggest that they are acting as
agents for the Government. I want to be really clear here. It
is just that the laws we have, once we collect this data----
Senator Nelson. But they could act--that information could,
but I do not want to get off on that tangent.
Let me ask you this, Ms. Harris. Do you believe consumers
are entitled to an opt-in arrangement rather than the opt-out?
Ms. Harris. I think it depends on the data and I think it
depends on the context.
Senator Nelson. Give me an example.
Ms. Harris. Well, I think that we think that when you are
operating in an environment where you are the ISP and you are
standing between the ends of the network, that we already have
law, the Electronic Communications Privacy law, Cable Act, and
a myriad of State laws, that require an opt-in. The ISP stands
as sort of the center of Internet chain of trust between the
two ends.
And so online in the advertising context, if we could
figure out a robust way to do opt-in, but then where we have
difficulty is separating PII and non-PII, personal information
is starting to merge. It is much harder to separate them. We
have profiles of what seemingly each piece looks like, you
know, an innocuous piece of information, and they are tied to
some kind of ID so they can be refreshed, and that ID can be
tied to your Internet address. It is a very complicated area.
That is why I am really pleased to hear Senator Dorgan talk
about more hearings. I think we have to get a baseline privacy
bill, but we are going to have to do a lot of work to go
through----
Senator Nelson. I want to commend Senator Dorgan for doing
this, and this is the beginning of a long road as we try to
grapple with this issue.
May I ask one more question?
Senator Dorgan. Yes. Your time is expired but we will
recognize you for one more question. We are going to have a
second round because we have until 11:45. So you may proceed.
Senator Nelson. Right now we have got the Commission doing
this self-regulatory principle and it deals with the internal
information security, but it may not address the security of
these third parties. I wanted to ask you, Ms. Parnes, has the
Commission studied the types of data security or encryption
associated with these activities?
Ms. Parnes. The Commission has not specifically studied
whether all of this information that is being collected is
encrypted, but one of the principles that we have announced
calls for data security so that the information that is
collected is held in a secure way. And we have also called for
information to be held only as long as it is needed for some
legitimate business or law enforcement purpose. We do not focus
particularly on encryption or any other type of technology in
looking at data security. It is reasonable security measures in
light of the nature of the information that is held.
Senator Nelson. Thank you.
Senator Dorgan. Senator Nelson, thank you very much.
Let me make a couple of comments then ask some additional
questions. To give you an example of, as a consumer, things
that I think are beneficial in terms of online collection of
information, I go to--maybe I should not use the site name--
Amazon as an example and am searching for a book or a couple
books. And Amazon comes back to me and says here is what you
searched for previously. So all of a sudden, I know they have
kept information about me. Correct? And then they also come
back and they say, and by the way, here are other books that
people are looking at when they look at the book you are--so
they are obviously collecting information about what other
people have done. Very beneficial and very interesting to me to
take a look at that. I do not walk away from that computer
thinking, you know, that is a real serious problem. I think
that is an advantage.
But then there are other questions. For example, Ms.
Horvath, Google. I use Google as a verb because I just google
it. I am always googling something. Right? And so is Senator
DeMint I will bet. And I also use MSN. So the two of you,
Google and MSN, likely have information about where I have been
traveling. Right?
I did not do this. So I go to--well, I may not google this,
but I may. I decide I am going to get to WebMD somehow or
another and I am going to take a look at gout, dementia, and
postnasal drip. Right? Now, I do not want the whole room to
know that, not that I would do that.
[Laughter.]
Senator Dorgan. But that may apply to some others, but I
disavow it.
But my point is I do that and then I say, I want to find a
flight to San Francisco. I want to go to the St. Francis Hotel,
and then I want to go to a Flying Other Brothers concert. So I
am doing all that.
So now both of you perhaps know all of that about me or
maybe that was all Senator DeMint that did that. So let me ask
you a question. Do you know that about me?
Ms. Horvath. Actually we do not. It would really depend on
how you are using our service.
Senator Dorgan. Well, let us say I use it so that you know
it.
Ms. Horvath. If you are signed-on account holder, if you
are signed in, then we would know what you are searching for,
but we know what you are doing only on the Google.com site.
Once you leave the Google.com site, the connection is gone. We
are not collecting----
Senator Dorgan. If I am on MSN and pull up Goggle on MSN
and use Google to go find something, you are saying that you do
not keep a record of that?
Ms. Horvath. We would know what your search query was.
Senator Dorgan. Right.
Ms. Horvath. If you came onto our service to Goggle.com and
you were not logged in, what we would collect would be your IP
address, the operating system that you are using, the browser
type that you are using, the query that you have requested, and
we would, if you do not have a correct cookie already----
Senator Dorgan. That was my point.
Ms. Horvath.--we would have a cookie.
Senator Dorgan. That was my point.
Ms. Horvath. And that is all we would collect.
Senator Dorgan. Well, that is a lot.
Ms. Horvath. If you went out to the airline site----
Senator Dorgan. You know about postnasal drip then. Right?
Ms. Horvath. No.
Senator Dorgan. You do not know that is where I went?
Ms. Horvath. But if you were searching on Google for
postnasal drip, we would know that you were searching for that,
but once you went off----
Senator Dorgan. Right. That is my point.
Ms. Horvath.--to WebMD, we would not know that you were
searching for gout or----
Senator Dorgan. So how long do you keep the postnasal drip
inquiry?
Ms. Horvath. Well, it is only connected to your IP address.
There is an argument as to whether IP addresses are personally
identifiable information because of the nonstatic nature. When
you are assigned an IP address, it does not mean that you are
going to have it 3 days from now, the same IP address. So it
would be stored in our logs iteratively. So it would not be
saying Senator Dorgan was looking for postnasal drip in our
logs. No, we would not know that.
Senator Dorgan. I am using a silly example, obviously.
Ms. Horvath. No, I understand.
Senator Dorgan. But if I am on MSN, as an example, and I am
moving around, and then I go to Google and type in dementia,
MSN has a record of what I am doing, I assume, and you have a
record of what I am doing at Google. My question is, how long
do you keep those records?
Ms. Horvath. We store our search logs for 18 months, and
after 18 months, we anonymize them by deleting what would be
your cookie ID and also by redacting the IP address so it could
no longer be connected to a computer that you used.
Senator Dorgan. And how long do you store this information?
Mr. Hintze. Ours is 18 months as well and our anonymization
process goes a little bit further. We delete the entirety of
the IP address and all cookie ID's.
Senator Dorgan. And if you are a Gmail user and log in,
then what?
Ms. Horvath. Then the logs are exactly the same. It is 18
months and then it is deleted after that.
Senator Dorgan. And if Mr. Dykes comes to you at Google or
MSN some day and says, I want to contract with you all, or goes
to Verizon or whomever, a service provider or an online
company, and says, I would like to contract, I would like to
get everything that you have got, just stream it over to me,
because he is an advertising agency and he is going to frame up
advertising in the future that will have beneficial content for
somebody, your reaction to that? He says, I can actually pay
you some pretty big money if you just stream all your stuff to
me at the same time you are collecting it.
Mr. Hintze. It is a hypothetical. We would, obviously, look
at all the privacy implications of any deal we would do.
Currently we are not sharing that data with anybody. It is all
used for our own internal operations. As far as I am aware,
that is our plan going forward.
Senator Dorgan. Yes, Mr. Dykes?
Mr. Dykes. I would point out that if we were involved in
that type of transaction, we would not be storing the sensitive
medical information that you just cited, and we would not have
the IP address because we do a one-way hash on that. So we
would just keep it against an anonymous identifier the
innocuous commercial categories that occurred there.
Senator Dorgan. I understand.
Ms. Harris, Senator DeMint, my colleague, indicated--and I
think Mr. Crews also--that this is probably a solution in
search of a problem, this discussion. And Mr. Crews' point, I
think is that if somebody is doing something you do not like,
you are going to go someplace else. If you back all the way up
from that, that is like you do not need FDA inspecting food
plants because if somebody is producing food that makes you
sick, they will be out of business soon. But I mean, some make
that point that are really against all regulation. But is this
a solution in search of a problem?
Ms. Harris. No, I do not think it is. You would have to
have a level of transparency. And believe me, we encourage
every day our colleagues at this table and some who are not and
industry to compete with each other on privacy. But I do not
think consumers understand at a level of granularity about the
differences in policies. And I can say you are still keeping
information far too long. We still have information tied to IP
logs. I mean, this question about everything being anonymous,
everything is not anonymous. At a minimum, we have got pseudo-
anonymous logs. We have got one-way hashes. Those identifiers
can be linked back by somebody to the IP address because we are
updating the profiles. So it is just not that simple.
Senator Dorgan. One further point and then I will call on
my colleague.
Mr. Dykes. Could I interject to say that clearly represents
a misunderstanding of how we operate? And I will take time to
talk to Ms. Harris some more.
Senator Dorgan. If we have time, we will come to that at
the end.
One final point. This Committee did, on behalf of
consumers, a Do-Not-Call List, obviously having to do with
telephone solicitation. I am assuming that there are technical
difficulties with a Do Not Track List. I have just described,
when I started my questioning, why tracking with respect to my
search for a book on a site really is probably beneficial to
me, but I do not know what is being tracked of Senator DeMint's
or my or Senator Carper's activities on the web. I do not have
the foggiest idea who is tracking it, how they are tracking it,
how they might use it, whether that company has some scruples
and might be very careful about how it handles it or whether it
is somebody else that grabs a hold of it and says, you know
what, Katie, bar the door, I will sell it to anybody.
So that is the only question in my mind, that there are so
many unanswered questions about information, how people
navigate this web. The purpose of hearings, of course, is to
try to inform and to understand, and that is the purpose of
this hearing.
Let me call on Senator Carper who has not yet had an
opportunity, and then I will call on Senator DeMint.
STATEMENT OF HON. THOMAS R. CARPER,
U.S. SENATOR FROM DELAWARE
Senator Carper. I apologize for missing your presentations.
As you know, we have a number of Subcommittees we serve on, and
I have been detained at another hearing trying to figure out
how to save the Medicare Trust Fund a couple hundred millions
of dollars before it goes defunct.
And we are going to start voting I think in about 10
minutes.
Let me just use this as an opportunity to ask each of our
witnesses to give me--maybe a minute apiece--a take-away. If we
cannot remember everything you say--I would not pretend to. But
each of you, give us a take-away. When we leave here, what
would you have us keep in mind? Ms. Harris, why do we not start
with you? Then we will just go to Mr. Kelly and Mr. Crews.
Ms. Harris. I mean, I think the key take-away is that we
are in a very robust online environment where we get great
benefit from the advertising, but we are collecting more and
more information about consumers at the time that the
information that consumers are putting online is increasingly
personal and sensitive, health data, location data. Self-
regulation is a piece, but self-regulation alone is not enough
to protect privacy. And we need to have some baseline rules in
place.
We have hearings and we bring in the companies who are good
actors, who are at least willing to talk to privacy groups. It
is a very big Internet, and we have to have some baseline for
everybody.
Senator Carper. And who should provide the baseline
guidance? Should the industry police itself?
Ms. Harris. Well, I am saying self-regulation plays a key.
It forces the boats up, but it is not in and of itself a
solution and we think that Congress has to pass a baseline
privacy law. We have been advocating for years for that.
Senator Carper. All right. Thank you.
Mr. Kelly?
Mr. Kelly. Facebook has been focused on transparency in the
collection of information about what people do online. Senator
Nelson offered an excellent example earlier. If he was reading
an article--let us say it was in the Miami Herald online, the
only way that Facebook would know that he is reading that
article is if he hits the share button on his browser and it
says, you know, share it with my Facebook friends. And at that
point, the information would be collected. There is real-time
notice. It would show up in his mini-feed of activity on the
site and be presented to him.
And we are also committed to--if somebody wanted to target
any advertising against that, if he wanted to become a fan of
the Miami Herald on Facebook, that would be done in an
anonymous space. The advertiser would not find out that Senator
Nelson was a fan of the Miami Herald. They would just be able
to target advertising toward fans of the Miami Herald on
Facebook.
Senator Carper. All right. Thank you.
Mr. Crews?
Mr. Dykes. So----
Senator Carper. Go ahead, Mr. Dykes. You had a point on
this?
Mr. Dykes. Yes. I was going to say NebuAd would welcome
regulation in this area that was technology- and business
process-neutral and focused on the privacy elements. Obviously,
there should be strong controls, for more sensitive information
and more personal information, including pseudo-anonymous
information as Ms. Harris mentioned. I think that would be the
criteria on which rules should be established. But room should
be left for innovation because we just do not know how the
Internet is going to evolve in the future, and so there is
definitely a role for self-regulation by industry groups in
this sphere as well.
Senator Carper. Does anybody on our panel agree with
anything that Mr. Dykes just said? I saw some nodding.
Ms. Harris. I came close.
[Laughter.]
Senator Carper. I think, Mr. Crews, it is your turn.
Mr. Crews. I was just going to mention that if the kind of
profiling that we are all worried about here today is an
inherent feature of the Internet, one of the things we have to
worry about is what criminal elements are doing. In this
debate, it is not just self-regulation or no regulation. There
is a lot of competitive discipline that has to take place and
will take place here.
But because of new technologies--as I said, it is only
2008. There are a lot more technologies that are going to come
to fruition, including biometrics, including the face scanners
and that data getting incorporated. There are a lot of
institutions that have to evolve. I mean, markets do not pop
into being, but the institutions that surround them to
legitimize what industries do and to legitimize and establish a
set of business practices that make sense that consumers live
with have to evolve along with those new technologies. You do
not want regulations that impede that.
I mean, take a look at the things that do work. Look at
where opt-out is working. Look at where companies are already
adopting opt-in procedures for certain kinds of information,
and pay close attention to how important that is that those
kinds of evolutions happen. Do not do something that thwarts
them because there could be political failures just as there
are market failures.
One of the inherent problems here too is that compulsory
databases that Government mandates of all sorts can get blurred
with private databases, and then that creates a problem. I call
it the Social Security problem sometimes. You know, one of the
trickiest personally identifiable indicators is the Social
Security number. Well, here we are with an Internet that has
come out. We cannot pull the rug out from under commerce and
ban the use of the Social Security number, but we can move
toward a future where business generates its own identifiers
apart from that.
So those kinds of evolutions have to occur and we do not
want to do specific regulations that might impede that because
it is not a matter of self-regulation or no regulation. It is
competitive discipline or----
Senator Carper. Thank you.
Is it Mr. Hintze or Mr. Hintze?
Mr. Hintze. Hintze.
Senator Carper. Hintze. OK. I was wrong on both counts. Mr.
Hintze.
Mr. Hintze. I think there are a couple of take-aways from
today.
One, we believe that protecting consumer privacy is not
only compatible with the business of online advertising, it is
essential to the success of the online advertising business. If
consumers lose trust in this, if lawmakers are uncomfortable
with how it is operating, the reaction could ultimately
undermine this very important business model.
Microsoft itself has taken a number of concrete steps to
protect consumer privacy that we think show our leadership in
this space, but we are a relatively small player in the online
advertising space. So we believe that also baseline privacy
legislation is appropriate, supplemented by robust self-
regulation, supplemented by technology tools, and supplemented
by consumer education. They all need to work together to
protect consumer privacy in the space.
Senator Carper. Thank you.
And Ms. Horvath?
Ms. Horvath. I would agree with what the other panelists
have said. I guess our key take-away would be the need for a
baseline Federal privacy statute which would be based upon the
fair information practices that have been around for 20 years
and would give consumers some feeling of accountability, that
the companies have accountability to them for what they say
they are doing with their information.
We would also support self-regulatory standards for the
advertising sector. We also are already supportive of the NAI
which has the first set of self-regulatory standards that are
actually in effect right now.
Senator Carper. Thank you.
Ms. Parnes, you are here, but I believe----
Senator Dorgan. Your time has expired.
Senator Carper. OK, thank you very much.
Senator Dorgan. Senator DeMint?
Senator DeMint. Thank you, Mr. Chairman. I know we are out
of time.
A last comment. I spent all my career in advertising. So I
know for many, many years we have been able to buy mailing
lists that reflect behaviors. If someone subscribes to an
outdoor magazine, you know they are likely to buy other things.
One of the exciting things about the Internet is,
recognizing that most of the jobs in our country are created
and maintained by small businesses, the Internet gives small
companies the opportunity to invest advertising in a very
targeted way. And the ability to use behavioral data to create
market baskets that would allow smaller companies to focus
their ad dollars on people who are most likely to buy their
products is very, very important. And it is important that we
maintain this. And I think it is very important that our
Government look at creating laws and not try to manage the
business, which is a very difficult line for us to draw.
My hope would be that the industry would recognize that
consumers need to know what data about them is being collected
and held, and that needs to be disclosed and very transparent
so that the consumers become the regulators of the Internet. If
they know, they will be able to switch to different content
providers, different servers that are doing it better and
better.
So my encouragement to the panel, those of you who are
involved in the industry, is not to ask the Federal Government
to come in at this point and to attempt to regulate because
that is much different than a law that says how you can use
data. I think as you continue to develop your technology, the
disclosure and transparency to consumers is the best way to
make sure that we end up with a vehicle that not only benefits
consumers but is great for our economy.
So, again, Mr. Chairman, I thank you and I think this has
been very eye-opening and enlightening. And I yield back.
Senator Dorgan. Senator DeMint, thank you very much.
Senator Carper, did you wish to ask additional questions?
Senator Carper. On a lighter note, I have just come from a
hearing where everyone spoke and testified to us using
acronyms. And one guy used five acronyms in one sentence, and I
had no idea what he said. I have no idea what he said.
[Laughter.]
Senator Carper. I walk in here and people are talking about
one-way hashes and cookies. I think the Chairman threw in
something I think called the Flying Other Brothers. I am pretty
good on music, but that is a new one to me. I was not here
long, but it certainly has been illuminating. I obviously have
to update my dictionary.
Thank you.
Senator Dorgan. Did you understand dementia and postnasal
drip?
[Laughter.]
Senator Carper. All too well.
[Laughter.]
Senator Dorgan. The Flying Other Brothers is actually a
band that includes one of the Grateful Dead members, and they
just have very bright tie-dyed T-shirts.
[Laughter.]
Senator Dorgan. Let me thank all of the witnesses for being
here. I intend to have a hearing asking a number of the
Internet service providers to come and visit with us. I think
one of the most important elements coming from this hearing is
how little we do understand. I think knowledge is important
here. So I think most of us would like to understand what we
can about what is happening, what kind of information is
collected about our habits, about our movements, about our
travels on the Internet. The Internet is really a wonderful
thing. It brings the world to your fingertips.
Because I come from a town of 250 or 280 people, in my high
school we had a coat closet-sized library. I mean, literally a
coat closet that they turned into a library, just a few books.
I do not need someone suggesting that is obvious.
[Laughter.]
Senator Dorgan. But I do think now in those little schools,
the libraries are accessible by the Internet, the best
libraries in the world. It really brings the world to our
fingertips.
But there are legitimate questions raised about our
traveling over the Internet and who watches us and how that
information is used. I think we need to understand much more
about it. There are those who raise questions about the use of
information. I think those are very important questions. The
reason that the Federal Trade Commission has developed a set of
guidelines, self-regulatory guidelines, is because I think you
understand the potential exists for abuse.
And I think the last thing that Senator DeMint described is
something I certainly agree with. I would hope that every
consumer has an opportunity, when traveling on the Internet, to
understand what kind of information trail they leave and who
might want to use that, or who is using it, or how is
information about what is happening or what they are doing on
the Internet going to be used. Information will give people an
opportunity to decide, do they like the policies of this
particular site or that particular provider, or do they wish to
move elsewhere where policies are, they feel, more beneficial
to their privacy.
So I think that this hearing has been instructive for us,
and we will be announcing another hearing in which we will
discuss this with the Internet service providers.
Senator Carper. Mr. Chairman, would you yield for just a
moment?
Senator Dorgan. Senator Carper?
Senator Carper. In terms of the consumers being empowered
to shop and to use the knowledge and understanding of these
policies, probably every week we receive in our mail policy
disclosures from a financial services company that we deal
with. And my guess is that most of us do not take the time to
look at it, and if we did, a lot of us would not understand it.
So it is important if we are going to really enable the
marketplace to work and to harness market forces effectively,
the information has to come in ways that consumers can actually
internalize it, understand it, and find useful.
So thank you.
Senator Dorgan. I was just thinking that I frequently--and
you perhaps do--get a letter in the mail from the North Dakota
Drivers License Bureau saying someone made an inquiry about my
driver's license and any potential infractions. I guess that is
part of being in politics. Fortunately, it is clean and has
nothing attached to the license. But that happens frequently. I
assume there are a lot of people out there wondering about my
driver's license. But at least we are notified. You are
notified. I am notified when someone makes an inquiry and
wishes to get that information.
And so that represents kind of the overhanging question
here about who wants information, how do they use that
information that represents personal information or other
information about your travels on the Internet.
Let me thank the witnesses again. I appreciate your
patience. And it works out that the vote, I believe, has just
started.
This hearing is adjourned.
[Whereupon, at 11:48 a.m., the hearing was adjourned.]
A P P E N D I X
State of Connecticut
Hartford, CT, July 9, 2008
Hon. Daniel K. Inouye,
Chairman,
Hon. Ted Stevens,
Vice-Chairman,
Senate Committee on Commerce, Science, and Transportation,
Washington, DC.
Dear Senators Inouye and Stevens:
I appreciate the Senate Committee on Commerce, Science, and
Transportation holding a hearing on the critically important issue of
tracking consumer Internet use for marketing purposes. I urge
expeditious Federal action to stop Internet service provider and third
party marketer tracking of consumer Internet use for marketing purposes
without prior and explicit consumer approval.
Monitoring consumer Internet browsing is a gross invasion of
privacy for the sake of profit. It threatens to make every consumer's
life an open book. Widely strewn to unknown websites and marketers
would be highly sensitive and personal information such as medical
conditions or family problems and financial interests.
In this brave new world, every movement or activity by consumers on
the Internet will be recorded, collected and compiled into huge
databases and then sold to marketers. Consumers will be bombarded with
relentless and repeated advertising. Their personal activities and
interests will be exposed to potential security breaches, just as
countless breaches nationwide have opened private confidential
financial information to potential misuse and identity theft.
This problem is neither speculative nor specious, It is very real
and it is imminent. We are on the cusp of a new deep, enduring
paradigm, fraught with perils to privacy.
Charter Communications, a cable and Internet service provider with
customers in Connecticut and throughout the nation, recently announced
a pilot program to give consumers ``an enhanced Internet experience.''
These so-called enhanced services amounted to nothing more than spying
on consumer web browsing by NebuAd. After I called on Charter to stop
this initiative, it announced that it was canceling the pilot testing
program. Charter has failed to disavow or deny future plans to track
consumer Internet activities. Two phone companies--Embarq Corp. and
CenturyTel--have completed trial tracking programs. Besides NebuAd,
other tracking marketers are seeking targets of opportunity.
While Congress has sought to protect consumer privacy by enacting
legislation such as the Cable Communications Policy Act, 47 U.S.C. 551
et seq., and the Electronic Communications Privacy Act, 18 U.S.C. 2510
et seq., both laws must be strengthened to emphatically and effectively
ban tracking by Internet service providers and third party marketers.
Congress should act promptly to address this new Internet menace.
I urge your quick and decisive action. I hope to be of assistance
to the Committee in its work on this important initiative. Thank you.
Very truly yours,
Richard Blumenthal,
Attorney General.
______
Response to Written Questions Submitted by Hon. Maria Cantwell to
Lydia B. Parnes
Question 1. Ms. Parnes, do you believe consumers read online
privacy policies?
Answer. Available research suggests that some consumers read online
privacy policies, but many do not. Moreover, those that do read online
privacy policies may not do so consistently, and they typically do not
review policies for changes after the initial read. For example, in a
2006 survey, only 20 percent of consumers said that they read the
privacy policy when first providing information to a website ``most of
the time,'' although 43 percent said they had read an online privacy
policy. Only 5 percent of consumers reported that they frequently check
to see if privacy statements have been updated or revised.\1\
---------------------------------------------------------------------------
\1\See TRUSTe/TNS Survey (December 2006), available at https://
www.truste.org/about/press_release/12_06_06.php.
---------------------------------------------------------------------------
Although the reasons that consumers do not read online privacy
policies may vary, we believe one key reason is that privacy policies
are often too difficult to understand and/or too long. Research
confirms this. For example, in a 2008 survey, the main reasons
consumers reported for not reading privacy policies were because the
polices contain too much legalese or jargon (57 percent) and that they
take too long to read (58 percent).\2\ Another study found that typical
privacy policies require college-level reading skills to understand.\3\
In addition, several studies show that significant numbers of consumers
believe that the mere presence of a privacy policy on a website
indicates some level of substantive privacy protection for their
personal information.\4\ As a result, once these consumers see that a
privacy policy exists, they may believe it unnecessary to read the
policy.\5\
---------------------------------------------------------------------------
\2\ See AOL/DMS, ``2008 AOL Consumer Survey on Behavioral
Advertising'' (February 2008).
\3\ Irene Pollach, ``What's Wrong with Online Privacy Policies?,''
Communications of the ACM, 50(9), 103-108 (2007).
\4\ See, e.g., 2007 Golden Bear Omnibus Survey (47.8 percent of
consumers surveyed believe that, if a website has a privacy policy, it
cannot share information with affiliate companies, and 55.4 percent
believe that it could not sell information to other companies); Joseph
Turow, Lauren Feldman, and Kimberly Meltzer, ``Open to Exploitation:
American Shoppers Online and Offline,'' Annenburg Public Policy Center,
University of Pennsylvania (June 2000) (59 percent of consumers
surveyed believe that a privacy policy on a website means that the site
will not share consumer information with other websites or companies).
\5\ See, e.g., Joseph Turow and Chris Jay Hoofnagle, ``The FTC and
Consumer Privacy In the Coming Decade,'' at 17, presented at
``Protecting Consumers in the Next Tech-Ade'' (Nov. 8, 2006), available
at http://works.bepress.com/chris_hoofnagle/4/.
---------------------------------------------------------------------------
The FTC staff recognized the concerns about online privacy policies
when it issued its Proposed Self-Regulatory Principles. Because online
behavioral advertising is largely invisible and unknown to consumers,
the principles recommend that companies provide greater transparency
about the practice through a ``clear, concise, consumer-friendly and
prominent'' disclosure--that is, not through a disclosure cloaked in
legalese and buried in a privacy policy. We are encouraging companies
to develop creative ways to provide this disclosure, including by
placing it outside of the privacy policy.
Question 2. Ms. Parnes, does the Commission view an End User
Licensing Agreement as a contract between the online consumer and the
website visited?
Answer. An ``End User License Agreement,'' ``Terms and Conditions''
page, or similar document available on a website may be an enforceable
contract between the online consumer and the website visited, depending
on the particular circumstances and applicable state contract law.
Courts have frequently held that when consumers click an ``I Agree''
icon in an online transaction (a.k.a., ``clickwrap agreements''), and
the terms and conditions to which they agree are readily available for
review in advance via a hyperlink or in a scrollable window, those
consumers are bound by those terms and conditions.\6\ It is less clear
whether courts would hold that an ordinary consumer would be bound by
an agreement or other document where the consumer uses the website
without actual knowledge of the terms and conditions posted there.\7\
---------------------------------------------------------------------------
\6\ See, e.g., Novak v. Overture Services, Inc., 309 F. Supp.2d
446, 450-51 (E.D.N.Y. 2004) (court holds that plaintiff, by clicking an
``I accept'' icon agreeing online to be bound by the ``Terms of
Service'' governing use of an online discussion group set forth in a
scrollable window, viewable ten lines at a time, was bound by the forum
selection contained therein).
\7\ See, e.g., Specht v. Netscape Commc'ns Corp., 306 F.3d 17, 23-
24, 35-38 (2d Cir. 2002) (refusing to enforce agreement against
consumers); see also Douglas v. Talk America Inc., 495 F.3d 1062, 1066
(9th Cir. 2007) (per curiam) (customer of long-distance service
provider not bound by new contractual terms requiring arbitration of
dispute when service provider merely posted them on its website and
gave no notice to customer); Waters v. Earthlink, Inc., 91 F.App'x 697,
698 (1st Cir. 2003) (refusing to enforce an arbitration clause posted
on a website in the absence of proof the consumer had seen the clause).
---------------------------------------------------------------------------
However, in enforcing Section 5 of the FTC Act, it is generally
unnecessary for the Commission to determine whether a EULA is, or is
not, a contract. Under core FTC deception principles, disclosures
buried in a EULA cannot be used to contradict a company's other
representations, nor are they adequate to correct misimpressions that
the company's representations would otherwise leave.\8\ The Commission
analyzes EULA-only disclosures on a case-by-case basis, weighing what
information is material to consumers and the overall, net impression of
the transaction.\9\
---------------------------------------------------------------------------
\8\ See, e.g., Zango, Inc., FTC Docket No. C-4186 (2007) (complaint
alleged that adware distributor represented that consumers could
download free software and games and failed to disclose adequately that
adware was bundled in the download; adware was often disclosed only in
lengthy terms and conditions or through inconspicuous hyperlinks); FTC
v. Odysseus Marketing, Inc., No. 1:05-CV-00330-SM (D. N.H. 2006)
(complaint alleged that defendants deceptively failed to disclose
adequately that their software would collect consumers' personal
information and substantially alter behavior of computers where those
functions were disclosed only in EULA accessible via inconspicuous
hyperlink).
\9\ See Sony BMG Music Entertainment, FTC Docket No. C-4195, Letter
responding to comment from Jerry Berman, Center for Democracy and
Technology (June 28, 2007); Zango, Inc., FTC Docket No. C-4186, Letter
responding to comment from Mark Bohannon, Software & Information
Industry Ass'n (Mar. 7, 2007).
Question 3. Ms. Parnes, the FTC focuses on consumer harms. Can you
give me a couple of examples of the types of potential consumer harms
from online behavioral advertising? Does the Commission view a loss of
any degree of personal privacy as being ``a harm?''
Answer. The greatest risk of harm arises when information collected
about a consumer's online activity is retained but not properly
protected. Here, the Commission's work on data security informs our
concern that information is sometimes retained beyond the point when it
is needed, or without being well-protected. One possible harm from the
collection of information for behavioral advertising is that it could
be hacked or otherwise obtained and used for unauthorized purposes.
Second, online behavioral advertising can lead to advertising and other
communication with a consumer that reveals highly personal information
and that can be unwelcome if it relates to sensitive issues, such as
health or children, or is delivered in a shared computer environment.
For example, in a situation where multiple users share the same
computer, the delivery of behaviorally-targeted advertising might
reveal the fact that a user conducted searches relating to AIDS,
domestic abuse, or sexual preference.
As to whether a loss of personal privacy constitutes a ``harm,'' it
depends on the consumer. Different consumers may have different
expectations about how their information is collected and used in this
context, as well as different views about the information that they are
willing to provide to obtain certain benefits--for example, coupons or
free services. Indeed, in a survey conducted by Alan Westin,\10\ 59
percent of the respondents were not comfortable with online tracking,
yet 41 percent were ``very'' or ``somewhat'' comfortable.'' This is the
main reason that Commission staff has proposed that companies provide
consumers with choice concerning the collection of their data for the
purpose of delivering behavioral advertising. Consumers who are
comfortable with the practice can allow it, and consumers who are not
comfortable can decline.
---------------------------------------------------------------------------
\10\ See Alan Westin, ``How Online Users Feel About Behavioral
Marketing and How Adoption of Privacy and Security Policies Could
Affect Their Feelings,'' at 3 (March 2008).
Question 4. Ms. Parnes, does the ability for advertisers to be able
to link through a common field online, anonymous, behavioral marketing
data and personally identifiable data typically used by marketers in
the brick and mortar world provide any specific challenges to the
Commission as it looks toward finalizing its principles on online
behavioral advertising?
Answer. The ability to link non-personally identifiable information
with personally identifiable information, and the debate concerning
what online information is personally identifiable and what is not,
have been among the central issues discussed in this area and in the
comments to our proposed principles. Especially as technology advances,
the line between the two categories of information becomes less and
less clear. To the extent that non-personally identifiable information
can become personally identifiable through reasonable technological
efforts or linkages with other data, we believe it raises a concern. We
are considering this issue carefully as we analyze the comments to our
proposed principles and consider next steps.
Question 5. Ms. Parnes, I realize the Commission's principles on
online behavioral advertising are just in a draft stage. In general,
does the Commission have the authority to enforce any principle for
self-regulation it may develop? If it does have this authority, how
does the Commission intend to enforce the self regulatory principles on
online behavioral advertising once they are finalized?
Answer. Several of the proposed self-regulatory principles reflect
requirements of existing law and the Commission has made enforcing
these principles a high priority. The agency will continue to enforce
laws within our jurisdiction as necessary to protect consumers. For
example, the principles maintain that companies should provide
reasonable security for behavioral data so that it does not fall into
the wrong hands. The Commission has brought numerous enforcement
actions focusing on the obligation of companies that collect or store
consumer data to provide reasonable security for that data.\11\ In
addition, the principles provide that before a company uses behavioral
data in a manner that is materially different from promises made when
the data was collected, it should obtain affirmative express consent
from the consumer. The Commission has brought high-profile law
enforcement actions against companies that violated this principle by
using data in a manner materially different from promises the company
made at the time of collection.\12\ In addition, if a company made
material misrepresentations about the collection or use of behavioral
advertising data, such misrepresentations would constitute a deceptive
practice, in violation of Section 5 of the FTC Act.
---------------------------------------------------------------------------
\11\ Since 2001, the Commission has obtained twenty consent orders
against companies that allegedly failed to provide reasonable
protections for sensitive consumer information. See In the Matter of
The TJX Companies, FTC File No. 072-3055 (Mar. 27, 2008, settlement
accepted for public comment); In the Matter of Reed Elsevier Inc. and
Seisint Inc., FTC File No. 052-3094 (Mar. 27, 2008, settlement accepted
for public comment); United States v. ValueClick Inc., No. CV08-01711
(C.D. Cal. Mar. 13, 2008); In the Matter of Goal Financial, LLC, FTC
Docket No. C-4216 (April 15, 2008); In the Matter of Life is Good,
Inc., FTC Docket No. C-4218 (Apr. 18, 2008); United States v. American
United Mortgage, No. CV07C 7064, (N.D. Ill. Dec. 18, 2007); In the
Matter of Guidance Software, Inc., FTC Docket No. C-4187 (Apr. 3,
2007); In the Matter of CardSystems Solutions, Inc., FTC Docket No. C-
4168 (Sept. 5, 2006); In the Matter of Nations Title Agency, Inc., FTC
Docket No. C-4161 (June 19, 2006); In the Matter of DSW, Inc., FTC
Docket No. C-4157 (Mar. 7, 2006); United States v. ChoicePoint, Inc.,
No: 106-CV-0198 (N.D. Ga. Feb. 15, 2006); In the Matter of Superior
Mortgage Corp., FTC Docket No. C-4153 (Dec. 14, 2005); In the Matter of
BJ's Wholesale Club, Inc., FTC Docket No. C-4148 (Sept. 20, 2005); In
the Matter of Nationwide Mortgage Group, Inc., FTC Docket No. 9319
(Apr. 12, 2005); In the Matter of Petco Animal Supplies, Inc., FTC
Docket No. C-4133 (Mar. 4, 2005); In the Matter of Sunbelt Lending
Services, FTC Docket No. C-4129 (Jan. 3, 2005); In the Matter of MTS
Inc., d/b/a Tower Records/Books/Video, FTC Docket No. C-4110 (May 28,
2004); In the Matter of Guess?, Inc., FTC Docket No. C-4091 (July 30,
2003); In the Matter of Microsoft Corp., FTC Docket No. C-4069 (Dec.
20, 2002); In the Matter of Eli Lilly & Co., FTC Docket No. C-4047 (May
8, 2002).
\12\ See, e.g., Gateway Learning Corp., Docket No. C-4120 (Sept.
10, 2004), http://www.ftc.gov/opa/2004/07/gateway.shtm.
---------------------------------------------------------------------------
The purpose of the FTC staff's proposed principles is to encourage
more meaningful and enforceable self-regulation. Because strong
enforcement mechanisms are necessary to ensure effective self-
regulation, it is our expectation that the organizations developing
self-regulation will include in their regimes mechanisms for meaningful
enforcement.
Question 6. Ms. Parnes, as you know, in recent months, there have
been deals announced between online search engines with strong online
advertising market shares. Considering the implications these proposals
have for market consolidation, is the Commission worried about the
prospect of the vast majority of behavioral online information being in
the hands of one company?
Answer. Market consolidation that results in the creation of
vaster, more detailed databases of online consumer data may raise
concerns. Although companies with large stores of data can be just as
privacy-protective as those with small ones, the risks associated with
data collection and storage, such as the risk that data could fall into
the wrong hands, may be heightened where one company maintains large
quantities of rich data. Further, competition ensures that companies
have incentives to protect customer privacy, and market consolidation
could undermine competition in this area.
However, whether there is cause for concern must be evaluated on a
case-by-case basis. A given deal between online advertising companies
may not involve the transfer or sharing of any data, or may not result
in a more detailed database of consumer information than those already
possessed by other companies in the online marketplace. The Commission
can evaluate whether a proposed transaction would adversely affect non-
price attributes of competition, such as consumer privacy, and will
continue to do so as appropriate.
______
Response to Written Questions Submitted by Hon. David Vitter to
Lydia B. Parnes
Question 1. Do you believe legislation mandating the FTC's pending
self-regulatory principles is necessary at this time?
Answer. I do not believe that legislation regarding online
behavioral advertising is necessary at this time. As stated in the
written testimony presented to the Committee on July 9, 2008,\1\ the
Commission is cautiously optimistic that the privacy issues raised by
online behavioral advertising can be addressed effectively through
self-regulation.
---------------------------------------------------------------------------
\1\ See Privacy Implications of Online Advertising: Hearing Before
the S. Comm. on Commerce, Science, and Transportation, 110th Cong.
(2008) (statement of the Federal Trade Commission), at 14.
---------------------------------------------------------------------------
The online marketplace is undergoing rapid change. Many different
types of businesses are entering the advertising market and the
technologies utilized by these businesses are constantly evolving. At
the same time, the costs and benefits of various types of behavioral
advertising may be difficult to weigh. Although online behavioral
advertising raises legitimate privacy concerns, it may provide benefits
to consumers in the form of more personalized and relevant
advertisements, as well as the free content that Internet users have
come to expect. In this environment, industry self-regulation, which
may afford the flexibility needed as online business models and
technologies evolve, may be the preferred approach. Although there is
much work to be done in this area, I believe that the self-regulatory
processes that have been initiated by the Network Advertising
Initiative (NAI) and other organizations should be given an opportunity
to develop.
Question 2. Do you believe the same principles about transparency
and choice would be necessary for behavioral advertising based on
``anonymous'' or non-personally-identifiable information?
Answer. The FTC received considerable comment on this issue in
response to the staff's proposed self-regulatory principles. Many
commenters stated, for example, that the collection of data that is not
personally identifiable is unlikely to cause consumer harm and
therefore should not be subject to the notice and choice requirements
in the staffs proposed principles. Other commenters stated that, even
when information collected is not personally identifiable, its use can
lead to advertising or other contacts with a consumer that can be
unwelcome or embarrassing, especially if they relate to sensitive
issues, such as health, children, or a consumer's finances, or are
delivered in a shared computer environment.
The comments also highlight the considerable debate that remains
concerning what online information is personally identifiable and what
is not, and the effect that advances in technology may have on the
distinction. Further, incidents such as the AOL breach demonstrate that
items of information that are considered anonymous standing alone may
become personally identifiable when combined, challenging traditional
notions about what data is or is not personally identifiable.\2\
---------------------------------------------------------------------------
\2\ Although AOL took steps to anonymize the search records of its
subscribers that were made public in 2006, several newspapers and
consumer groups were able, to identify some individual AOL users and
their queries by combining the data.
---------------------------------------------------------------------------
FTC staff continues to carefully review the comments received on
this issue, and we intend to address it as we develop our next steps in
this area.
Question 3. Based on your experience, how is online behavioral
advertising different from offline direct marketing? Also, what would
be the justification for regulating online advertising differently than
offline advertising?
Answer. Targeted marketing also exists in the offline environment.
However, rapidly evolving Internet technologies permit online marketing
companies to collect, store, and analyze significantly greater amounts
of data than offline companies, and to do so through practices that are
often invisible to consumers. Offline marketing generally involves the
collection of a smaller quantity of less detailed behavioral
information than is gathered online, and that collection is likely to
be more consistent with consumer expectations.
For example, brick-and-mortar stores often record a consumer's
purchase history and use it to market to the consumer. They typically
do not, however, follow individual consumers around their stores and
collect detailed information about the other products the consumer
might have looked at, nor do they share purchase or behavioral
information with third parties. Similarly, consumers know that
subscribing to a newspaper involves the sharing of personal information
so that the newspaper can be delivered to their homes; however, they
generally would not expect every article read in a newspaper to be
tracked and recorded by multiple parties. Given the amount and type of
data that is collected and stored in connection with online behavioral
advertising, and the invisibility and typically unexpected nature of
such practices, it may be appropriate to take different approaches to
protecting consumer privacy between online behavioral advertising and
offline marketing.
Question 4. Based on your experience, has the FTC identified any
specific harm to consumers that has resulted from the information used
for online behavioral targeting?
Answer. At this time, the FTC is unaware of any specific incidents
in which a company's online behavioral advertising practices have led
to such consumer harms as identity theft, financial fraud, or physical
harm. Behavioral advertising, however, does raise unique concerns
because of the amount, richness, and sensitivity of the consumer
information that may be collected, used, and shared with third parties.
For example, in a situation where multiple users share the same
computer, the delivery of behaviorally-targeted advertising might
reveal the fact that a user conducted searches relating to highly
personal and sensitive topics, such as AIDS, domestic abuse, or sexual
orientation. To the extent that the information collected is or could
later become personally identifiable, the risk of harm may be greater.
Many consumers express concern about the privacy and data security
implications of behavioral tracking. For example, a 2008 survey showed
that 59 percent of those surveyed are ``not very comfortable'' (34
percent) or ``not comfortable at all'' (25 percent) with websites using
online activity data to tailor advertisements or content to their
hobbies and interests, whereas 41 percent were ``comfortable'' or
``somewhat comfortable.'' \3\ As this survey indicates, consumers
differ in their level of concern about this practice and in how they
weigh the benefits of more relevant ads against the privacy
implications. For this reason, the Commission staff's proposed
principles would require that companies provide consumers with choice
concerning the collection of their data for the purpose of delivering
behavioral advertising.
---------------------------------------------------------------------------
\3\ See Alan Westin, ``How Online Users Feel About Behavioral
Marketing and How Adoption of Privacy and Security Policies Could
Affect Their Feelings,'' at 3 (March 2008).
---------------------------------------------------------------------------
______
Response to Written Questions Submitted by Hon. Maria Cantwell to
Chris Kelly
Question 1. Mr. Kelly, as you mentioned in your written statement,
Facebook's privacy policy states clearly that you believe all of your
users ``should have control over [their] personal information.''
However, as I understand it from my staff, many of whom have Facebook
accounts, when members of Facebook wish to terminate the service, they
are only allowed to ``deactivate.'' Do you currently offer your members
an opportunity to control their personal information by completely
terminating their account and allowing Facebook to then delete their
personal information? If not, why hold onto their information?
Answer. At Facebook, we reflect our desire to give users control
over their information by providing them two options if they wish to
suspend or terminate their relationship with us.
Facebook users can make either a temporary or permanent decision to
shut down their account--the former is called deactivation, and the
latter deletion. Users choose to deactivate their accounts for a
variety of reasons. Given Facebook's roots as a college site, one
original driver for introducing the deactivation feature was user
desire to ``disappear'' during their exams and then resume the service
with their friend connections and network membership intact. More than
50 percent of the users who deactivate their Facebook accounts return
to reactivate them within weeks.
Deletion is a fuller option that erases the personally identifiable
information such as name, e-mail address, IM handle and other core
information from a Facebook account. While we cannot certify that every
single piece of data a user has ever given us is irretrievable after
the deletion process is done--the distributed nature of databases and
Internet site operations, especially for a longstanding user, makes
that certification practically impossible--we have scrubbed our active
databases of all contact information for a particular account. A
deleted account cannot be reconstructed; a user who wants to come back
to use the Facebook service after undergoing the deletion process must
start over with regard to setting up their friends and networks. I am
unaware of a similar deletion option offered by any other major
Internet company.
Question 2. Mr. Kelly, according to a Washington Post article from
last month, Facebook requires 95 percent of its members that use one of
the social network's 24,000 applications to give the applications'
developers access to their personal online profile, except contact
information, and their friends' profiles as well. While I understand
these developers are not allowed to share with advertisers, this still
leaves the 400,000 developers with access to personal information that
is out of Facebook's control. A recent study at the University of
Virginia found that about 90 percent of the most popular Facebook
applications have unnecessary access to private data. So my question
is, why does a Sudoku puzzle have to know that someone has two kids?
And why does a book-sharing program need to know my birthday?
Answer. Facebook's developer terms of use do not just forbid
sharing with advertisers of the information that applications request.
They only allow retention of most data called from the Facebook service
for 24 hours in order to facilitate the more rapid operation of the
application. Retention for a period beyond that is forbidden. Requests
for data through the Facebook API are logged, and the platform policy
enforcement team conducts investigations as necessary to discover
potential violations. They then take action up to and including barring
an application or a developer from further use of the service where
violations are discovered.
Users of course may choose to establish a deeper relationship with
an application by providing more personal information directly, but
Facebook's terms and policies work together to encourage strongly
clarity with users about access to their data.
In addition to the technical and policy enforcement measures
outlined here, Facebook is always looking for means to enhance
transparency to users with regard to data collection and use. We are
currently exploring efficient and effective means to give users greater
knowledge of and control over data requests by applications so that a
user will know if an application is seeking more data than the user
believes is necessary. At that point, the user could then choose to
remove the application from their profile, or to use the ``block''
feature that has been present since the introduction of the Facebook
platform in 2007 that prevents any data from flowing to that
application.
______
Response to Written Questions Submitted by Hon. David Vitter to
Chris Kelly
Question 1. Does your company's business model already accommodate
the FTC's proposed principles for industry self-regulation? If so,
please explain how.
Answer. In putting in place baseline privacy controls that users
can adjust to their liking and allowing users to make their own choices
about the sharing of personal information with advertisers, we
generally reflect the FTC self-regulatory principles. Our commentary on
the principles has suggested that the FTC offer greater clarity in
distinguishing between how personally and non-personally identifiable
information should be addressed by the principles, which we believe
will lead to greater consumer understanding and confidence.
Question 2. Does your system accommodate for a consumer's choice
not to receive behavioral advertising, and in your systems, is that
request honored permanently? If so, please explain how.
Answer. While Facebook does not currently engage in many of the
behavioral targeting practices that have been the main focus of the
Committee's attention, it bears noting at the outset that Facebook is
an opt-in system at its core. If a user does not sign up for and use
Facebook, they will not receive advertising through our service.
In addition to this fundamental user choice, we offer opt-outs for
many of our advertising products within the Facebook ecosystem. For
instance, those users who do not wish to participate in our Social Ads
product have an easily available option to turn it off.
User control is a critical part of Facebook's philosophy and our
offerings in the product and advertising area will continue to reflect
that principle.
______
Response to Written Questions Submitted by Hon. Maria Cantwell to
Jane Horvath
Question 1. Ms. Horvath, just a few months ago, in Google's
comments to the FTC on self-regulatory principles for online
advertising, the company said that ``contextual advertising . . . is
not the type of advertising that ought to be the focus of the FTC's
efforts to develop effective self regulatory principles.'' I understand
that contextual advertising is less invasive than behavioral
advertising. However, I am interested to hear the rationale behind
Google's belief that it should be exempt from the FTC's self-regulatory
principle.
Answer. Google has not suggested that it or any other online
advertiser be exempted from self-regulation with respect to behavioral
advertising. We have said and continue to believe that further thought
must be given to the definition of behavioral advertising in the
Federal Trade Commission staff's draft self-regulatory principles. As
currently written, the draft principles define ``behavioral
advertising'' so broadly as to encompass virtually any collection and
use of information about a user's online activities, including the
collection and use of information to provide contextual advertising.
AOL, Google, Microsoft, Yahoo!, and many other companies provide
contextual advertising solutions, which, as you point out, are
different from behaviorally targeted ads because they provide relevant
advertising based on what an Internet user is searching for as well as
relevant ads based on a page that a user is viewing. Though it is not
the focus of our business today, Google believes that behavioral
advertising can be done in ways that are responsible and protective of
consumer privacy and the security of consumers' information.
To ensure the continuation and proliferation of responsible
behavioral advertising practices, we are supportive of efforts to
establish strong self-regulatory principles for online advertising that
involves the collection of user data for the purpose of creating
behavioral and demographic profiles. For example, we believe that the
FTC's efforts to address this type of advertising through self-
regulatory principles are appropriate and helpful. Likewise, we support
the Network Advertising Initiative's recently-announced draft Self-
Regulatory Code of Conduct for Online Behavioral Advertising, which
includes limitation on the use of sensitive information to create
profiles of individuals for purposes of third-party advertising.
For both the FTC's draft principles and the NAI's draft code of
conduct, we believe that the focus on data collected across multiple
web domains owned or operated by different entities to categorize
likely consumer interest segments for use in online advertising is
appropriate. We also believe that a strong and easy-to-find mechanism
to permit consumers to opt out of this type of data collection is a
goal that all companies should aspire to achieve. Finally, we believe
that special attention should be given to rules around the creation of
profiles based on sensitive information such as health status.
Question 2. Ms. Horvath, one of the central issues in the Internet
privacy debate is the protection, or lack thereof, of personally
identifiable information. I'm interested to know, what measures Google
is taking to ensure that the data it uses for ad targeting is not
connected to personally identifiable information? What opt-out
mechanisms do you currently offer your customers? How robust are these
opt-out mechanisms and how easy are these mechanisms for consumers to
use?
Answer. Google protects its users' personally identifiable
information--PII--in many ways. For example, we do not use PII to serve
ads in connection with our AdWords and AdSense products. We also have
strict policies and procedures in place to ensure that personal
information is used only in accordance with our privacy policy, which
is located at www.google.com/privacy.html, and that our users' personal
information is secure.
We collect non-PII through the DoubleClick cookie in order to
enable enhanced functionality to advertisers that use DoubleClick and
advertisers that advertise on the Google content network through our
AdSense for Publishers product. For example, this data collection will
enable advertisers that advertise through AdSense for Publishers to
limit the number of times a user sees an ad through frequency capping.
Users will have a better experience on Google content network sites
because they will no longer see the same ad over and over again.
Users are able to opt out of data collection through our
DoubleClick ad serving cookie in several ways. For example, users can
opt out by visiting the DoubleClick opt out page located at
www.doubleclick.com/privacy. In addition, our ads privacy microsite has
an above-the-fold opt out button located at www.google.com/
privacy_ads.html. Users are also able to opt out of the DoubleClick
cookie's data collection through the Network Advertising Initiative's
opt out page located at www.networkadvertising.org/managing/
opt_out.asp. This single opt-out is honored both in the DoubleClick
network and in the Google content network.
Question 3. Ms. Horvath, in your written testimony, you indicate
Google is currently experimenting to deliver search results based on
prior searches. Isn't this behavioral advertising by any other name? Is
this consistent with Google's privacy policy regarding behavioral
advertising?
Answer. We are currently experimenting with providing ads on Google
search based on both a user's current query and his or her recent prior
queries. For example, a user who types ``Italy vacation'' into the
Google search box might see ads about Tuscany or affordable flights to
Rome. If the user were to subsequently search for ``weather,'' we might
assume that there is a link between ``Italy vacation'' and ``weather''
and deliver ads regarding local weather conditions in Italy. In the
above example, we are serving an ad based on a user's activity on our
site, and not the sites of other parties. This is an example of the
kind of first-party advertising that users expect to see in response to
the search terms they enter into the Google search box. While we may
use recent queries to better respond to the user's specific interests,
this is done on the fly, and we are not building profiles based on a
users' search activities.
Google's privacy policy, which is located at www.google.com/
privacy.html, states clearly that we use this information to provide
our services to users, including the display of customized content and
advertising. We also provide plain English explanation of our
advertising and privacy practices on our ads privacy microsite located
at www.google.com/privacy_ads.html.
Question 4. Ms. Horvath, with regard to Google's recently announced
marketing deal with Yahoo!, are you able to tell us what information
Yahoo! will share with Google? Will it include IP addresses? If so, has
Google done its due diligence to ensure that the transfer of data from
Yahoo won't violate Yahoo!'s privacy policy? As you know, Yahoo! has
publicly announced it will retain information about consumer search
queries for 13 months. I believe your company retains such information
for 18 months. Does Google intend to conform to Yahoo!'s retention
policy or is Yahoo! expected to conform to Google's?
Answer. Under our advertising agreement with Yahoo!, there is no
PII passed between Yahoo! and Google. In fact, for ads appearing
(impressions) on Yahoo! Search, Yahoo! will remove the last octet of
the IP address before sending us search queries, so Yahoo will never
send the full IP address to Google. Moreover, in the search advertising
context Google will set a cookie only if a user clicks on an ad
delivered by Google.
______
Response to Written Questions Submitted by Hon. David Vitter to
Jane Horvath
Question 1. Does your company's business model already accommodate
the FTC's proposed principles for industry self-regulation? If so,
please explain how.
Answer. We have participated actively in the Federal Trade
Commission's efforts to develop privacy principles relating to online
privacy and behavioral advertising. Our hope is that the FTC's privacy
principles--once finalized and written to ensure that they can be
realized by industry and will provide consumers with appropriate levels
of transparency, choice, and security--will be adopted widely by the
online advertising industry and will serve as a model for industry
self-regulation in jurisdictions beyond the United States.
Our comments on the FTC's principles are available at
googlepublicpolicy.
blogspot.com/2008/04/our-comments-on-ftcs-behavioral.html. In addition,
through DoubleClick we are members of and serve on the Board of
Directors of the Network Advertising Initiative and abide by the NAI
Self-Regulatory Principles.
Question 2. Does your system accommodate for a consumer's choice
not to receive behavioral advertising, and in your systems, is that
request honored permanently? If so, please explain how.
Answer. Though our business is focused on contextual advertising,
we collect non-personally identifiable information through our
DoubleClick ad serving cookie in order to enable enhanced functionality
to advertisers that use DoubleClick and advertisers that advertise on
the Google content network through our AdSense for Publishers product.
For example, this data collection will enable advertisers that
advertise through AdSense for Publishers to limit the number of times a
user sees an ad through frequency capping. Users will have a better
experience on Google content network sites because they will no longer
see the same ad over and over again.
Users are able to opt out of data collection through our
DoubleClick ad serving cookie in several ways. For example, users can
opt out by visiting the DoubleClick opt out page located at
www.doubleclick.com/privacy. In addition, our ads privacy microsite has
an above-the-fold opt out button located at www.google.com/
privacy_ads.html. Users are also able to opt out of the DoubleClick
cookie's data collection through the Network Advertising Initiative's
opt out page located at www.networkadvertising.org/managing/
opt_out.asp. This single opt-out is honored both in the DoubleClick
network and in the Google content network.
We believe that a strong and easy-to-find mechanism to permit
consumers to opt out of this type of data collection is a goal that all
companies should aspire to achieve.
______
Response to Written Questions Submitted by Hon. Maria Cantwell to
Leslie Harris
Question 1. Ms. Harris, do you believe consumers read online
privacy policies?
Answer. There is ample evidence that consumers do not read privacy
policies and to the extent that they do, they do not understand them.
Indeed, a recent study found that over half of respondents believed
that the presence of a privacy policy on a website meant that the site
did not share or sell personal information. Moreover, privacy policies
often are in practice disclosure policies, making no promises to
protect customer information and instead stating all the ways the site
intends to use and disclose data. Consumers need clear, conspicuous and
understandable notice of collection and disclosure practices.
Question 2. Ms. Harris, do you view an End User Licensing Agreement
as a contract between the online consumer and the website visited?
Answer. End User Licensing Agreements are typically included when
consumers purchase or download software. Courts generally have held
that these EULAs do constitute contracts between consumers and the
companies distributing the software, though it may depend on factors
such as the prominence of the EULA and whether consumers affirmatively
express agreement (by clicking an ``I Agree'' button, for example). As
a practical matter, as we have seen in the case of malicious
``spyware'' and as the FTC has recognized in its settlement agreements
with spyware purveyors, long, complicated EULAs are wholly inadequate
to inform consumers about the material effects that a particular piece
of software will have on their computers. The FTC has set the right
precedent in requiring simple explanations of such key information
outside of any software EULA. Outside the software context, it is less
clear that merely visiting a website would be sufficient to
contractually bind a consumer to a Terms of Service agreement,
particularly in the absence of some express indication of consumer
acceptance. The FTC has made, clear, however, that websites must adhere
to their publicly stated policies on matters such as privacy, or risk
possible charges of deception.
Question 3. Ms. Harris, as you know the FTC focuses on consumer
harms. Do you believe that some of the issues surrounding online
behavioral advertising go beyond what the Commission has traditional
viewed as consumer harm?
Answer. The FTC has traditionally focused on tangible harms such as
practices which result in a financial loss or which induce a consumer
to engage in a transaction through deceptive practices. We do not
believe that the privacy concerns raised by behavioral advertising are
limited to those which result in an adverse action against a consumer.
The right of privacy is harmed when consumers have no reasonable
expectation that information about them is being collected, they are
not provided meaningful notice that would allow them to gauge the
privacy risks, and they have no way to make meaningful decisions about
whether and how their information may be used.
Question 4. Ms. Harris, should online contextual advertising be
exempt from any self regulatory framework or does the Commission only
need to look at behavioral advertising?
Answer. Contextual advertising, which is often used to generate ads
alongside search results, matches advertisements to the content of the
page that a consumer is currently viewing. The privacy risks associated
with contextual advertising vary. If the practice is transparent to the
user and data collection and retention is minimal, the practice poses
little risk to the consumer. For example, if a site only looks at the
consumer's activity and only serves ads based on that activity
contemporaneously , i.e., at the moment the consumer is engaging in the
activity and does not collect and save the consumer data, the privacy
concern is low. But the privacy concerns are heightened if the user
data is retained in an identifiable or pseudonymous form (i.e., linked
to a user identifier) for long periods of time even if it is not
immediately used to create advertising profiles.
CDT has long advocated for a baseline privacy bill that would cover
all collection and use of consumer data and require the full range of
fair information practices, which include but are not limited to robust
notice and consumer choice but also provide penalties for noncompliance
and remedies for consumers. Self regulation is not enough. To the
extent that self-regulation remains as one component of privacy
protection in this area, consumers will certainly benefit from a self-
regulatory scheme that covers both contextual and behavioral
advertising, with escalating protections for models with increased data
collection, retention, use, and identifiability.
Question 5. Ms. Harris, are you concerned with the recently
announced marketing deal between Google and Yahoo! that one company
will have under its control the vast majority of consumers' online
behavioral information?
Answer. The Google/Yahoo! marketing deal does have the potential to
consolidate more data about search and search advertising under one
roof. While this data is certainly one component of behavioral
information, Google does not currently use search data to create
behavioral profiles. Thus, the more relevant question is whether the
consolidation of search information--not necessarily behavioral
information--is of concern.
This deal follows in the footsteps of a series of major mergers and
acquisitions in the online advertising space over the past 18 months.
All of this market consolidation means that more and more data about
what consumers do online is housed by fewer companies, exacerbating
existing privacy concerns about how such data is collected, used,
safeguarded, and shared. We believe that recent market dynamics are
even further evidence of the need for a general privacy law to protect
consumer data at large.
______
Question for the Record from Hon. David Vitter to
Leslie Harris
Question. As our committee continues to examine the issue of online
privacy, should we focus on the variations in different technologies
used to provide what appears to be essentially similar marketing
services? Or, should we instead focus on the use of the information
collected--by whatever method it is collected--to ensure that data is
used for legitimate marketing purposes, that privacy is protected, and
that we can go after those who misuse any data they collect?
[The witness did not respond.]
______
Response to Written Questions Submitted by Hon. Maria Cantwell to
Robert R. Dykes
Question 1. Mr. Dykes, over the last few weeks numerous Internet
service providers, including CenturyTel, Charter and Wide Open West
have either ceased using your service or shelved any plans they had to
do so. What does this tell you about your proposed business model and
consumer reaction to it?
Answer. NebuAd and its Internet service provider (``ISP'') partners
have always been and continue to be committed to the core privacy
principles of transparency and consumer control regarding NebuAd's
services. We support our ISP partners' decisions to delay their
implementation plans so they can ensure that in deploying NebuAd's
services, customers receive clear, direct, and prior notice that
NebuAd's services will take effect, thereby allowing subscribers to
make an informed choice regarding whether to participate.
NebuAd plans to continue to explain our existing business model and
better educate the public about the state-of-the-art privacy
protections that have been built-in to NebuAd's services, and, equally
as important, the process that we have established to continuously
improve on them. In support of this process to continuously advance
privacy protections beyond industry standards, NebuAd looks forward to
a continued open dialogue with legislators, regulators, and the
advocacy community.
NebuAd remains committed to delivering strong value to advertisers,
publishers, and ISPs while setting the gold standard for privacy in
online advertising and delivering the best Internet experience possible
to consumers.
Question la. Do you believe that consumers have the perception that
your technology will allow ISPs to watch over every move they make on
the Internet?
Answer. A close examination of NebuAd's technology, operations, and
privacy protections demonstrates that it is a responsible, privacy
conscious business. Unfortunately, some recent public statements from
various sources have presented inaccuracies, distortions, and
misrepresentations of our technology and business model. While some
consumers may have formed mistaken perceptions based on these erroneous
statements about the NebuAd technology, the privacy controls in place,
and the business model in conjunction with our ISP partners, we believe
that consumer education is a key component of our continued effort to
set the gold standard for privacy in online advertising.
Finally, as a point of clarification, NebuAd does not watch every
move consumers make on the Internet. The NebuAd system only uses a
select set of a consumer's Internet activities (that is, only a subset
of HTTP traffic) to construct anonymous inferences about the consumer's
level of qualification for a predefined set of market segment
categories, which are then used to select and serve the most relevant
advertisements to that consumer. The NebuAd system does not collect or
use any information from password-protected sites (e.g., HTTPS
traffic), web mail, e-mail, instant messages, or VoIP traffic, and the
system does not make use of any market segments that are deemed to be
sensitive.
Question 2. Mr. Dykes, in your written testimony you praise online
advertising, especially NebuAd's approach, for enhancing consumers'
Internet experience. As you know, the most important quality to
Internet users is the speed of their Internet connection. How does the
NebuAd's approach to behavioral monitoring--deep packet inspection--
affect the speed of a consumer's connection, including uploads and
downloads?
Answer. NebuAd's service does not adversely affect either upload or
download performance. The NebuAd Ultra Transparent Appliance (``UTA'')
is transparent when it comes to performance impact. Lab-tests of the
NebuAd UTA and have found the performance of the NebuAd UTA matched or
exceeded standard performance metrics expected from any network device
such as a switch and/or a router. In addition, NebuAd has sophisticated
monitoring capabilities to ensure performance expectations are
maintained.
______
Response to Written Questions Submitted by Hon. David Vitter to
Robert R. Dykes
Question 1. Does your company's business model already accommodate
the FTC's proposed principles for industry self-regulation? If so,
please explain how.
Answer. NebuAd participated in the FTC's proceeding by submitting
written comments on the proposed principles. A copy of NebuAd's
comments is available at http://www.ftc.gov/os/comments/
behavioraladprinciples/080411nebuad.pdf (``NebuAd Comments''). In its
written comments, NebuAd agreed with the FTC Staff that the ``self-
regulatory principles that emerge from this process must rest within an
overall framework that promotes transparency, consumer control, limited
use of sensitive information, limited data storage, and strong
security.''
NebuAd's comments focused on a central theme: any set of final
proposals for self-regulation should focus on the ultimate goal--
preventing consumer harm--and not on regulating different behavioral
advertising technologies and companies in different ways based simply
on the underlying technology used or individual entities involved. In
other words, NebuAd believes that the final self-regulatory principles
that emerge from the FTC proceeding must be consistent with the twin
objectives of technology-neutrality and provider and publisher-
neutrality. This will allow for innovation within a flexible self-
regulatory framework while preventing an unintended consequence of
inadvertently picking winners and losers in the emerging behavioral
advertising marketplace, based solely on technology or business model.
NebuAd looks forward to the Commission Staff's release of its final
principles. Like the Federal Trade Commission, NebuAd is ``cautiously
optimistic'' that industry self-regulation will work to protect
consumers. See Prepared Statement of the Federal Trade Commission on
Behavioral Advertising, Before the Senate Committee on Commerce,
Science, and Transportation (July 9, 2008) at 1, available at http://
www.ftc.gov/os/2008/07/P085400behavioralad.pdf. Moreover, NebuAd has
joined a chorus of other companies in calling for baseline privacy
legislation. See oral testimony of Mr. Robert R. Dykes, Chairman and
CEO, NebuAd, Inc, in hearing before the Subcommittee on
Telecommunications and the Internet: What Your Broadband Provider Knows
About Your Web Use: Deep Packet Inspection and Communications Laws and
Policies, July 17, 2008.
NebuAd's specific response to this question is as follows. (The
FTC's Proposed Principles for the Self-Regulation of Behavioral
Advertising are available at http://www.ftc.gov/os/2007/12/
P859900stmt.pdf).
Principle 1: To address the need for greater transparency and
consumer control regarding privacy issues raised by behavioral
advertising, the FTC staff proposes:
Every website where data is collected for behavioral
advertising should provide a clear, consumer-friendly, and
prominent statement that data is being collected to provide ads
targeted to the consumer and give consumers the ability to
choose whether or not to have their information collected for
such purpose.
NebuAd Response: NebuAd supports the underlying goal of this
proposal, which is to ensure that consumers are provided with ample
opportunity to understand what information is collected, by whom, for
what purpose, and to provide consumers with control over their online
experience. Unfortunately, the proposal, as written, is problematic for
NebuAd for reasons that are business-model specific, and having nothing
to do with providing consumers with meaningful transparency and
consumer control. As we noted in our comments to the FTC,
``[t]his proposal is apparently addressed to behavioral
advertising companies that, unlike NebuAd, have direct
relationships with those websites, typically as part of a
network. Because NebuAd works through ISPs and other ad
networks and does not necessarily have direct relationships
with the websites consumers visit, it has no way to require
them to post the proposed notice. For this reason, NebuAd
respectfully asks the Commission Staff to consider alternative
methods of notice, such as direct notice provided by ISPs to
their subscribers, together with an opportunity to opt-out, as
an appropriate means of meeting the Staff's proposed
transparency principle.''
See NebuAd comments at p. 4. In other words, NebuAd is in full
agreement with the policy of this principle, but must meet it in an
alternative way. Indeed, that alternative way--direct notice to
consumers, prior to the NebuAd service taking effect, with an
opportunity to opt out then and persistently thereafter--may be more
transparent than the Commission's proposed transparency principle.
Principle 2: To address the concern that data collected for
behavioral advertising may find its way into the hands of criminals or
other wrongdoers, and concerns about the length of time companies are
retaining consumer data, the FTC staff proposes:
Any company that collects or stores consumer data for
behavioral advertising should provide reasonable security for
that data and should retain data only as long as is necessary
to fulfill a legitimate business or law enforcement need.
NebuAd Response: NebuAd is in complete agreement with this
principle. While NebuAd does not collect personally-identifiable
information and does not store raw data associated with individual
users, NebuAd nevertheless provides state-of-the-art security for the
data it does collect and store: commercial categories mapped against
anonymous identifiers and some aggregate data used for analytics.
Moreover, unlike some search some companies, NebuAd retains the data
used for behavioral advertising purposes only for as long as it is
useful for that purpose--generally from a few days to a couple of
months.
Principle 3: To address the concern that companies may not keep
their privacy promises when they change their privacy policies, FTC
staff proposes:
Companies should obtain affirmative express consent from
affected consumers before using data in a manner materially
different from promises the company made when it collected the
data.
NebuAd Response: NebuAd requires by contract that its ISP partners
not only change their privacy policies to address NebuAd's service, but
also to provide direct notice to subscribers at least 30 days prior to
the service taking effect, with an opportunity to opt-out. This allows
for subscribers to exercise informed choice prior to the service taking
effect. It is important to note that opting out does not mean that the
subscriber must find a new ISP; rather, the subscriber may stay with
the same ISP, without the NebuAd behavioral advertising service
offering more relevant ads than those the subscriber would otherwise
receive.
Principle 4: To address the concern that sensitive data--medical
information or children's activities online, for example--may be used
in behavioral advertising, FTC staff proposes:
Companies should only collect sensitive data for behavioral
advertising if they obtain affirmative express consent from the
consumer to receive such advertising.
NebuAd Response: The NebuAd system does not make use of any market
segments that are deemed to be sensitive. Specifically, NebuAd does not
track or serve ads based on visits related to adult content, sensitive
medical information, racial or ethnic origins, religious beliefs or
content of a sexual nature, and does not have market segment categories
for illegal products. Accordingly, the NebuAd system does not have nor
does the system ever attempt to detect keyword patterns related to such
subjects. NebuAd therefore looks forward to the final set of the FTC
Staff's ``sensitive'' categories, which NebuAd assumes will be
carefully tailored to prevent consumer harm, and, assuming so, will
exclude those categories from the data used for behavioral advertising
purposes or require an affirmative opt-in for them.
Question 2. Does your system accommodate for a consumer's choice
not to receive behavioral advertising, and in your systems, is that
request honored permanently? If so, please explain how.
Answer. Yes. NebuAd's service and policies were designed to provide
consumers with prior, robust notice and the opportunity to express
informed choice about whether to participate, both before the service
takes effect and persistently thereafter.
If a consumer should exercise his/her choice to not receive
behavioral advertising, the consumer's choice is honored in the same
manner as all other typical ad networks that provide ``cookie-based
opt-out mechanisms'' and consistent with the self-regulatory guidelines
of the Network Advertising Initiative (``NAI'').
In addition, as stated on the opt-out page displayed to the
consumer (either on the NebuAd website or on our partner ISP website),
we inform that consumer that if he/she should delete the NebuAd cookie
or if he/she should change computers or web browsers, then the consumer
would need to opt-out again. This is similar to the statement that the
members of the NAI make to consumers (http://
lwww.networkadvertising.org/managing/opt out.asp).
NebuAd's current opt-out system is a more robust mechanism than
traditional ``cookie-based opt-out'' systems and goes beyond the system
offered by typical ad networks in the following situation. If a
consumer's web browser blocks cookies, the NebuAd system will consider
the consumer to be an opted-out user and will exclude that consumer
from NebuAd's information collection and targeted ads.
Finally, NebuAd recently announced that it is developing a network-
based opt-out mechanism that is not reliant on web browser cookies.
Leveraging this advanced technology, ISP partners will be empowered to
offer this enhanced mechanism to their subscribers in order to honor
their opt-out choices more persistently than current systems widely
used today.
Question 3. Can you expand on how NebuAd's model collects
information for marketing, depersonalizes that information into market
segments, and keeps it depersonalized? To be clear, I am not asking
that you reveal any of your company's proprietary information, only a
general overview of how this is done and how consumers' information is
protected in your system.
Answer. The NebuAd system only uses a select set of a consumer's
Internet activities (that is, only a subset of HTTP traffic) to
construct anonymous inferences about the consumer's level of
qualification for a predefined set of market segment categories
(``anonymous user profiles''), which are then used to select and serve
the most relevant advertisements to that consumer.
Depersonalization into Market Segments
An anonymous user profile is a set of numbers that represent the
consumer's level of qualification for a predefined set of NebuAd
supported market segments (e.g., Las Vegas travel or domestic SUVs).
NebuAd develops a set of keyword patterns associated with each specific
market segment.
As HTTP traffic flows through, the NebuAd system looks for
appearances of these keyword patterns. A consumer's level of
qualification for each particular market segment increases for each
detected keyword pattern appearance. None of a consumer's HTTP traffic
or the keyword patterns themselves is ever stored within an anonymous
user profile. Only the set of numbers that represent the consumer's
level of qualification, at a given point in time, for a limited number
of broad market segments is maintained within an anonymous user
profile. This mechanism of constructing anonymous inferences about a
consumer's level of qualification and not storing the raw data that was
used to create or update a user's anonymous profile provides a strong
additional layer of privacy protection that goes beyond the standards
used by many Internet companies today.
In addition, each market segment has a predefined lifespan
associated with it so if no keyword pattern for that market segment is
detected for some time, the consumer's level of qualification will age
fairly quickly--generally from a few days to a couple of months.
Specifically, this means that a consumer's anonymous user profile
represents just his/her current qualification levels and does not
retain a long-standing history of qualifications levels.
Additional Protections
The NebuAd system has also built-in multiple additional privacy
protections from the ground up to ensure a consumer's anonymity
including:
The NebuAd system does not collect or use any information
from password-protected sites (e.g., HTTPS traffic), web mail,
e-mail, instant messages, or VoIP traffic.
As noted above, the NebuAd system does not make use of any
market segments that are deemed to be sensitive. Specifically,
NebuAd does not track or serve ads based on visits related to
adult content, sensitive medical information, racial or ethnic
origins, religious beliefs or content of a sexual nature, and
does not have market segment categories for illegal products.
Accordingly, the NebuAd system does not have nor does the
system ever attempt to detect keyword patterns related to such
subjects.
Finally, by design, the NebuAd system's set of keyword
patterns do not contain any personally identifiable information
about Internet consumers, and it ensures the anonymous
information that its systems infer cannot be used to identify
any individual. None of the anonymous information NebuAd uses
can be compiled together and somehow reverse engineered to
identify any individual. In other words, the information is not
``pseudo-anonymous."
______
Memorandum
July 8, 2008
From: Nebuad, Inc.
Re: Legal and Policy Issues Supporting Nebuad's Services
I. Introduction to NebuAd
NebuAd is an online media company founded by Internet security
experts in 2006. It provides online advertising in partnership with
ISPs, using a select set of a user's Internet activities (only a subset
of HTTP traffic) to construct anonymous inferences about the user's
level of qualification with respect to a predefined set of market
segment categories (``anonymous user profiles''), which are then used
to select and serve the most relevant advertisements to that user.
NebuAd is a newcomer to the world of online advertising. This world
of Internet companies includes several industry giants, behavioral
advertising networks, and countless website publishers. Currently,
online advertising solutions operate in many locations throughout the
Internet ecosystem--from users' computers to individual websites to
networks of websites. When an Internet user visits the sites of web
publishers, like Yahoo! or Amazon, these sites typically collect
information about the user's activities to target ads based on that
information. When an Internet user conducts a search, the search
company may collect information from the user's activity, which in turn
may be used to improve the relevance of the sponsored search results
and ads shown. When a user visits websites within an online advertising
network, some of which include thousands of sites, the visits help the
advertising network track the user for the purpose of serving higher-
value targeted advertising. All of these activities are well-entrenched
in the Internet and have become fundamental to the economic model that
underpins the wide availability of content and services on the Internet
today. These advertising capabilities, have proven to have mutual
benefits for users, publishers--both large and small--and advertisers.
NebuAd offers a unique business model that allows ISPs to
participate in the online advertising ecosystem, while not only
adhering to industry-standard privacy policies but also establishing
new state-of-the-art privacy protections and user choice policies that
go far and beyond those used on the Internet today.
Given the background of its founders, NebuAd architected its
service and its policies to adhere to very strict privacy principles.
These include:
1. NebuAd's service does not collect or use PII from ISP
subscribers. The entire ad optimization and serving system does
not collect or use any Personally Identifiable Information
(PII), nor does it collect any information from password-
protected sites, web mail, e-mail, instant messages, or VOIP
traffic.
2. NebuAd stores only a set of numbers that represent the
user's level of qualification for a predefined set of market
segment categories (``anonymous user profiles''). NebuAd does
not store raw data such as URLs navigated or IP addresses
associated with an identifiable individual.
Rather, the NebuAd service constructs anonymous inferences
about the user's level of qualification for a predefined set of
market segment categories, and then discards the raw data that
was used to create or update a user's anonymous profile. This
mechanism of constructing anonymous inferences about the user's
level of qualification and not storing raw data provides a
strong additional layer of privacy protection that goes beyond
the standards used by many Internet companies today.
3. NebuAd's ISP Partners are required to provide notice to
users in advance of launch of the service. The notice, which
must be direct and robust, discloses to the user that the ISP
is working to ensure that advertisements shown will be more
relevant advertisements, that to deliver these ads its partner
creates anonymous profiles based on part of the user's web
surfing behavior, which does not include the collection of PII,
and that the user may opt-out of the service. For existing
subscribers, the notice is required to be delivered 30 days
prior to the launch of the service by postal mail, e-mail, or
both. For new subscribers, the notice is required to be placed
clearly and conspicuously in the new subscriber sign-up flow
and outside the privacy policy. All subscribers can opt-out at
any time, and on-going disclosure and opportunity to opt-out is
required to be provided within the ISP's privacy policy.
4. NebuAd and its ISP partners offer users advance and on-going
choice of opting-out of the service. Users are provided with a
clear statement of what the opt-out means and the way it
operates. Once the opt-out option is chosen, NebuAd honors that
choice and ignores the user's subsequent web surfing activity
and thus does not serve the user with behaviorally targeted
ads.\1\
---------------------------------------------------------------------------
\1\ The user, of course, will continue to receive ads.
5. NebuAd's service only creates anonymous user profiles, which
contain no PII and no raw data, and its placement of ads is
completely anonymous. NebuAd uses proprietary algorithms and
techniques, including one-way encryption of data, so that no
one--not even NebuAd's engineers who designed the system--can
reverse-engineer an anonymous identifier, or the anonymous user
---------------------------------------------------------------------------
profile associated with it, to an identifiable individual.
6. NebuAd avoids any sensitive websites or product categories.
NebuAd does not track or serve ads based on visits related to
adult content, sensitive medical information, racial or ethnic
origins, religious beliefs or content of a sexual nature, and
does not have market segment categories for illegal products.
7. NebuAd does not permit either complexity of data or
narrowness of data to be reverse-engineered into PII. This
protection is accomplished because anonymous user profiles are
constructed by anonymous inferences about the user's level of
qualification for a predefined set of market segment
categories. Raw data is simply not stored as part of the
anonymous user profile. In addition, the NebuAd service does
not have narrowly-defined segments. Finally, the anonymous
profile identifier is the result of multiple encryptions, and
based on multiple data elements including the hashed IP
address.
8. There is no connection or link between the ISP's
registration data systems and NebuAd. That means that no user-
specific data is exchanged between NebuAd and ISP data systems.
This boundary is preserved further, and inadvertent disclosure
is prevented, because NebuAd immediately performs a one-way
encryption of the IP address and other anonymous user
identifiers used within the NebuAd system.
9. NebuAd installs no applications of any type on users'
computers, has no access to users' hard drives, and has no
access to secure transactions. As such, NebuAd does not control
a user's computer or web-surfing activity in any way, e.g., by
changing computer settings or observing private or sensitive
information.
10. NebuAd's Data Centers are professionally operated and
secured. NebuAd's servers are located at secure sites with
state-of-the-art protections against any intrusion, electronic
or physical.
II. The Federal Wiretap Act
As a threshold matter, it is important to note that the Federal
Wiretap Act \2\ was last amended in 1986 before the widespread adoption
of personal computing and online communications.\3\ When the Wiretap
Act was enacted, and amended, the focus was on telephone communication
and other similar technology. Case law is rich with examples of claims
involving a tapped phone line.\4\ Notably, these cases primarily
involve direct, one-on-one communication between the parties. The
content is personal to the speakers, such that if one of the parties
was replaced, the communication would not contain the same content.
Although secrecy or confidentiality was not expressly built into the
Wiretap Act, the Act was enacted at a time when the focus was on
individual communications--likely as a result of the limitations of
then-existing technology.
---------------------------------------------------------------------------
\2\ 18 U.S.C. 2510 et seq.
\3\ The Wiretap Act was amended by the Electronic Communications
Privacy Act of 1986 (``ECPA''), Pub. L. 99-508, 100 Stat. 1848 (1986).
While the Wiretap Act is Title I of the ECPA, it was first passed as
Title III of the Omnibus Crime Control and Safe Streets Act of 1968 and
is generally known as ``Title III.''
\4\ See, e.g., United States v. Foster, 580 F.2d 388 (10th Cir.
1978) (telephone company taps phone line of user suspected of
defrauding the telephone company out of long-distance charges); United
States v. Harvey, 540 F.2d 1345 (8th Cir. 1976) (same); United States
v. Auler, 539 F.2d 642 (7th Cir. 1976) (same).
---------------------------------------------------------------------------
The environment that has since evolved for online communications is
markedly different. While online communications are still carried by
wire, there are important policy distinctions between the types of
communications that the Wiretap Act was enacted to address, and the
types of communications present in the online environment today.
Internet users are not engaged in a personal, direct conversation with
non-secure website publishers.\5\ Such publishers provide online
content indiscriminately to all users. As stated below, even under the
Wiretap Act, courts look to the circumstances surrounding a
communication.\6\ Yet, the evaluation of circumstances that surround a
telephone communication between two parties is not analogous to an
online communication between a party and a website. To date, there are
no litigated decisions directly addressing the application of the
Wiretap Act to a URL provided as part of a consumer's online
navigations or provided via publicly available search request and
response. Therefore, it is still an open question as to whether these
types of communications are even covered by the Wiretap Act.\7\
---------------------------------------------------------------------------
\5\ There are always exceptions to this statement, such as online
purchases, encrypted communication, and other secured data
transactions, but notably, these private communications are the exact
types of information that NebuAd's services do not collect. NebuAd's
services personalize generic content rather than intruding upon private
communications.
\6\ See United States v. Amen, 831 F.2d 373, 378 (2d Cir. 1987).
\7\ See Patricia L. Bellia, Spyware: The Latest Cyber-Regulatory
Challenge, 20 Berkeley Tech. L.J. 1283, 1296, 1311-12 (2005). Another
law review article described the question as to whether URLs contain
contents as ``surprisingly difficult'' and ``quite murky.'' Orin S.
Kerr, Internet Surveillance Law After the USA Patriot Act: The Big
Brother That Isn't, 97 Nw. U.L. Rev. 607, 645-46 (2003).
---------------------------------------------------------------------------
Assuming, for the purposes of this memorandum, that the Wiretap Act
applies to NebuAd's services, the Act expressly prohibits the
intentional interception of an electronic communication \8\ unless
``one of the parties to the communication has given prior consent to
such interception.'' \9\ The legislative history of the Wiretap Act
clearly indicates ``that Congress intended the consent requirement to
be construed broadly.'' \10\ As a result, ``courts have resoundingly
recognized the doctrine of implied consent.'' \11\ The Court of Appeals
for the Second Circuit stated that the Wiretap Act ``affords safe
harbor not only for persons who intercept calls with the explicit
consent of a conversant but also for those who do so after receiving
implied consent.'' \12\
---------------------------------------------------------------------------
\8\ 18 U.S.C. 2511(1)(a).
\9\Id. 2511(2)(d).
\10\ United States v. Amen, 831 F.2d 373, 378 (2d Cir. 1987) (``
`Consent may be expressed or implied. Surveillance devices in banks or
apartment houses for institutional or personal protection would be
impliedly consented to.' '' (quoting S. Rep. No. 1097, 90th Cong. 2d
Sess., reprinted in 1968 U.S.C.C.A.N. 2112, 2182)).
\11\ George v. Carusone, 849 F. Supp. 159, 164 (D. Conn. 1994); see
United States v. Faulkner, 439 F.3d 1221, 1224-25 (10th Cir. 2006)
(``We are not persuaded to depart from the unanimous view of the
holdings by our fellow circuit courts.''); Griggs-Ryan v. Smith, 904
F.2d 112, 118; United States v. Corona-Chavez, 328 F.3d 974, 978-79
(8th Cir. 2003); Amen, 831 F.2d at 378; United States v. Willoughby,
860 F.2d 15, 19-20; United States v. Tzakis, 736 F.2d 867, 870, 872 (2d
Cir. 1984); Borninski v. Williamson, No. Civ. A. 3:02CV1014-L, 2005 WL
1206872, at *13 (N.D. Tex. May 17, 2005); United States v. Rittweger,
258 F. Supp. 2d 345, 354 (S.D.N.Y. 2003); Chance v. Avenue A, Inc., 165
F. Supp. 2d 1153, 1162 (W.D. Wash. 2001); In re DoubleClick Inc.
Privacy Litig., 154 F. Supp. 2d 497, 514 (S.D.N.Y. 2001).
\12\ Griggs-Ryan, 904 F.2d at 116.
---------------------------------------------------------------------------
To determine whether a party has impliedly consented to an
interception under the Wiretap Act, courts examine the totality of the
circumstances and ``imply consent in fact from surrounding
circumstances indicating that the [party] knowingly agreed to the
surveillance.'' \13\ In such evaluations, courts have found that
parties impliedly consented to an interception in various fact
patterns. The Federal district court for the Southern District of New
York found implied consent when an employer circulated memoranda
regarding telephone monitoring and recording. Although the party denied
receiving the notice, and evidence proving such receipt was destroyed,
the court determined that the party had knowledge of the monitoring and
recording and impliedly consented to such monitoring and recording by
continuing to use the monitored telephone lines.\14\
---------------------------------------------------------------------------
\13\ Amen, 831 F.2d at 378.
\14\ Rittweger, 258 F. Supp. 2d at 354.
---------------------------------------------------------------------------
Similarly, a Connecticut Federal district court found that
employees had given their implied consented to the recording of
conversations on work telephones, as many of the telephones displayed
warning labels, memoranda were circulated to all employees regarding
the recordingof incoming and outgoing telephone calls.\15\ The court
stated that employees' ``knowledge of the system and subsequent use of
the phones is tantamount to implied consent to the interception of
their conversations.'' \16\ The Court of Appeals for the First Circuit
held that repeated oral statements that all incoming telephone calls
would be monitored was sufficient notice, and that the party's taking
an incoming phone call was implied consent to the interception.\17\
Additionally, a Texas Federal district court found that an employee
consented to monitoring of Internet communications at work because the
employee had signed a form stating that ``Internet access should be
limited to `business use only,' and that the company `logs and archives
all incoming and outgoing data communications through its gateway
system. Use of the gateway implies consent to such monitoring.' '' \18\
---------------------------------------------------------------------------
\15\ George, 849 F. Supp. at 164.
\16\ Id.
\17\ Griggs-Ryan, 904 F.2d at 117-19.
\18\ Borninski v. Williamson, No. Civ. A. 3:02CV1014-L, 2005 WL
1206872, at *13 (N.D. Tex. May 17, 2005).
---------------------------------------------------------------------------
Using the framework established by the courts, NebuAd satisfies the
implied consent exception to liability for interception under the
Federal Wiretap Act.\19\ NebuAd requires, by contract, that all of its
ISP partners give subscribers notice of NebuAd's services, including
the collection of anonymous information regarding subscribers' online
activities, for use in advertising. This notice must be given directly,
and prior to the initiation of the ISP's use of NebuAd's services. The
ISP partners are also required, by contract, to alter their privacy
policies accordingly. NebuAd further requires that all ISP partners
provide users with an option to opt-out of NebuAd's services, initially
upon receipt of the direct notice, and in an ongoing manner through the
ISP's privacy policy.
---------------------------------------------------------------------------
\19\ Website publishers may also consent to an interception, as
website publishers make web content available for any user. Such
posting does not constitute an exclusive communication between the
website publisher and the user, but rather it is public communication
that is intended to be viewed by any number of simultaneous users. As a
result, website publishers have no reasonable expectation that the
communication between it and any consumer will remain private or
confidential, and thus impliedly consent to the interception by a third
party.
---------------------------------------------------------------------------
III. The Cable Act
The Cable Act \20\ was enacted to protect cable subscribers'
personal information. Among other things, it requires cable operators
to obtain written or electronic consent from a subscriber prior to
collecting any PII concerning the subscriber.\21\ In addition to the
limitations on the collection of subscriber PII, the Cable Act limits
the disclosure of subscriber PII by cable operators.\22\ The Cable Act
sets out multiple standards that a cable operator must satisfy in order
to disclose subscriber PII. If the disclosure is necessary for a
legitimate business activity, a cable operator is not required to
provide the subscriber with any notice.\23\ A cable operator may
disclose the name and mailing addresses of subscribers if it provides
subscribers with the opportunity to opt out of such disclosure.\24\ For
all other disclosures of subscriber PII, a cable operator must obtain
``the prior written or electronic consent of the subscriber''--
essentially an opt-in standard.\25\
---------------------------------------------------------------------------
\20\ Cable Communications Policy Act (1984), 47 U.S.C. 551 et
seq.
\21\ Id. 551(b)(1).
\22\ Id. 551(c).
\23\ Id. 551(c)(2)(A).
\24\ Id. 551(c)(2)(C)(i).
\25\ Id. 551(c)(1).
---------------------------------------------------------------------------
Notably, under the Cable Act, PII ``does not include any record of
aggregate data which does not identify particular persons.'' \26\
NebuAd's service specifically complies with the Cable Act because
NebuAd's service does not collect PII. Instead, using only non-
personally identifiable information, NebuAd uses a select set of a
user's Internet activities (a subset of HTTP traffic) to construct
anonymous inferences about the user's level of qualification for a
predefined set of market segment categories, which are then used to
select and serve the most relevant advertisements to that user. The use
of NebuAd's services certainly does not require a subscriber to opt
in--the strictest notice and consent requirement. Although not an
activity conducted by NebuAd, even the disclosure of a subscriber's
mailing address, widely recognized as PII, only requires that the
subscriber have an opportunity to opt out. NebuAd's service, on the
other hand, does not even collect subscriber PII. Because NebuAd's
service does not collect subscriber PII, there is no violation of the
Cable Act.
---------------------------------------------------------------------------
\26\ Id. 551(a)(2)(A).
---------------------------------------------------------------------------
Additionally, a 2002 FCC ruling concluded that ``cable modem
service, as it is currently offered, is properly classified as an
interstate information service, not as a cable service, and that there
is no separate offering of a telecommunications service.'' \27\ This
determination that cable Internet services are not classified as
telecommunications services was upheld by the Supreme Court as a lawful
interpretation of the Communications Act.\28\ A recent decision by the
Court of Appeals for the Sixth Circuit upheld this distinction and
stated that the plain language of the Cable Act precludes its
application to broadband Internet services, even those provided by a
cable operator.\29\ Examining the application of the Cable Act, the
court emphasized that as the cable provider was providing broadband
Internet access and not cable service, the Cable Act was inapplicable.
---------------------------------------------------------------------------
\27\ In re Inquiry Concerning High-Speed Access to the Internet
Over Cable and Other Facilities, 17 FCCR 4798, 4802 (2002).
\28\ Nat'l Cable and Telecomms. Ass'n v. Brand X Internet Servs.,
545 U.S. 967 (2005).
\29\ Klimas v. Comcast Cable Commc'ns, Inc., 465 F.3d 271 (6th Cir.
2007), reh'g en banc denied Klimas v. Comcast Cable Commc'ns, Inc.,
2007 U.S. App. LEXIS 13658 (6th Cir. May 1, 2007).
---------------------------------------------------------------------------
IV. Policy Implications
NebuAd provides users with a great amount of privacy protection.
Unlike many online advertising models today, NebuAd's service does not
collect or use any PII. In addition, NebuAd's anonymous user profiles
do not contain any original raw data, such as URLs navigated, but only
consist of a set of numbers that represent anonymous inferences about
the user's level of qualification for a predefined set of market
segment categories. (NebuAd does retain some anonymous data for
analysis and reporting.) Additionally, NebuAd is one of the only
models--if not the only model--that provides users with advance notice
of the nature of its services and an opportunity to opt-out before the
service takes effect. NebuAd's service also complies with the
government's consent policy on privacy as NebuAd's service does not
collect any PII, and provides users with the opportunity to opt-
out.\30\ Finally, NebuAd's service does not observe encrypted traffic,
does not observe VoIP sessions, does not store raw search queries
linked to an identifiable user, and does not track users' IP addresses,
thus providing an excellent set of privacy protections. Because of the
privacy protections that NebuAd has incorporated into the architecture
of its service, it is able to provide users with relevant advertising
messages in a safe, secure, and privacy-respecting manner.
---------------------------------------------------------------------------
\30\ Use of a consumer opt out is consistent with other consumer
information protection statutes such as the Gramm-Leach-Bliley Act
(financial data), the Health Insurance Portability And Accountability
Act (health data), the Fair Credit Reporting Act (consumer reports),
the Telemarketing and Consumer Fraud and Abuse Prevention Act
(telemarketing), and the CAN-SPAM Act (e-mail marketing).
---------------------------------------------------------------------------
______
Memorandum Addendum
July 8, 2008
From: Nebuad, Inc.
Re: Legal and Policy Issues Supporting Nebuad's Services
I. The Memorandum on Behavioral Advertising \1\ by the Center For
Democracy and Technology Is Based on a Misunderstanding of
Nebuad's Services
---------------------------------------------------------------------------
\1\ Center for Democracy and Technology, Privacy Implications of
Online Advertising (July 9, 2008) [hereinafter ``CDT Memorandum''].
---------------------------------------------------------------------------
NebuAd's service was architected to comply with very strict privacy
principles. The Center for Democracy and Technology (``CDT'')
misunderstands how NebuAd's service operates. First, the service does
not ``cop[y] all or substantially all Web transactions.'' \2\ NebuAd's
service uses only a subset of HTTP traffic to construct anonymous
inferences about the user's level of qualification with respect to a
predefined set of market segment categories. NebuAd's service only
stores a one-way encrypted anonymous user identifier, which is used to
represent an anonymous user, and a set of numbers which represent the
user's level of qualification with respect to a predefined set of
market segment categories. NebuAd does not store raw data such as URLs
navigated or IP addresses associated with an identifiable individual.
Second, to provide additional privacy protection, NebuAd's service does
not track or serve ads based on visits to sensitive websites or product
categories. CDT also unfortunately erroneously stated in its Memorandum
that a NebuAd ISP implementation ``did not provide a way for
subscribers to give or withhold consent.'' \3\ This is not so. NebuAd
requires its ISP partners, by contract, to give their ISP subscribers
prior, direct notice about NebuAd's service and an opportunity to
withhold consent or to express their informed choice before the service
takes effect.
---------------------------------------------------------------------------
\2\ CDT Memorandum at 23.
\3\ Id. at 15.
---------------------------------------------------------------------------
II. The CDT's Evaluation Is Largely a Policy Argument on How the
Existing Law Ought to Apply in the Internet Context
A. Federal Wiretap Act
While citing a number of cases and policy arguments, the CDT
acknowledges that exceptions to the Federal Wiretap Act may apply. The
CDT Memorandum contains hedging language that an exception ``probably
does not permit'' NebuAd's service,\4\ that it is ``unlikely'' that the
``necessary incident'' exception would apply,\5\ and citing no cases on
point, that it is ``unclear'' whether the ``business use'' exception
would apply.\6\ It is therefore unclear from the CDT's own Memorandum
whether NebuAd's service qualifies for either of these exceptions.
---------------------------------------------------------------------------
\4\ Id. at 26.
\5\ Id. at 26.
\6\ Id. at 27.
---------------------------------------------------------------------------
The CDT also does not deny that implied consent, where a consumer
receives notice, prior to the service taking effect, has the
opportunity to opt-out, and continues to use the service, can and often
amounts to implied consent under the Wiretap Act, which has been well-
recognized by the Federal courts. The CDT Memorandum equivocates on
NebuAd's service in particular, stating that prior notice of NebuAd's
service ``might not be enough,'' \7\ to meet the implied consent
standard, and that the courts would be skeptical ``if'' the notice to
consumers did not provide sufficient notice of the services.\8\
---------------------------------------------------------------------------
\7\ Id. at 30-31.
\8\ Id. at 31.
---------------------------------------------------------------------------
The CDT's Memorandum does not give a full picture of the implied
consent standard. While courts have stated ``that consent under the
Wiretap Act `is not to be cavalierly implied,' '' \9\ courts have
equally stated that ``Congress intended the consent requirement to be
construed broadly.'' \10\ These statements are not mutually exclusive,
and courts have made both statements in the same case.\11\
---------------------------------------------------------------------------
\9\ Id. At 30 (quoting Watkins v. L.M. Berry & Co., 704 F.2d 577,
579 (11th Cir. 1983).
\10\ Griggs-Ryan v. Smith, 904 F.2d 112, 116 (1st Cir. 1990); see,
e.g., United States v. Amen, 831 F.2d 373, 378 (2d Cir. 1987); United
States v. Faulkner, 439 F.3d 1221, 1224 (10th Cir. 2006); George v.
Carusone, 849 F. Supp. 159, 164 (D. Conn. 1994).
\11\ See, e.g., Griggs-Ryan v. Smith, 904 F.2d at 116-17.
---------------------------------------------------------------------------
Moreover, none of the cases the CDT relies on for its policy
argument that opt-in should be required is on point, and some contain
language affirming the use of implied consent. For example, reliance on
Watkins \12\ for the proposition that implied consent is insufficient
is misplaced, as Watkins involves an employee that had consented to
limited monitoring practices by an employer, and the employer
subsequently exceeded the authorized monitoring. Thus, Watkins does not
state that implied consent is invalid, but rather that consent may be
limited.\13\ Similarly, the CDT cites to Griggs-Ryan v. Smith \14\ for
the proposition that ``consent should not casually be inferred.'' The
context surrounding this statement is important in order to gauge the
statement's full meaning. In the preceding paragraphs of the same case,
the court described how ``Congress intended the consent requirement to
be construed broadly,'' and ``that Title III affords safe harbor not
only for persons who intercept calls with the explicit consent of a
conversant but also for those who do so after receiving implied
consent.'' \15\ The Griggs-Ryan court found implied consent based on
repeated oral statements that all incoming calls would be monitored.
The other cases CDT cites on implied consent are either distinguishable
or don't apply here.\16\
---------------------------------------------------------------------------
\12\ Watkins v. L.M. Berry & Co., 704 F.2d 577 (11th Cir. 1983).
\13\ See also Griggs-Ryan, 904 F.2d at 119 (finding the plaintiff's
heavy reliance on Watkins to be mislaid because of the limited consent
in that case).
\14\ Griggs-Ryan, 904 F.2d 112 (1st Cir. 1990).
\15\ Id. at 116.
\16\ See Pharmatrak, Inc. Privacy Litig., 329 F.3d 9, 19-20 (1st
Cir. 2003) (The court found that ``consent must be actual, not
constructive.'' The court in Pharmatrak indeed made this point but only
insofar as to cite to the decision from which it originated, namely,
Williams v. Poulos, 11 F.3d 271 (1st Cir. 1993).); United States v.
Corona-Chavez, 328 F.3d 974, 978 (8th Cir. 2003) (The court held that
there must be actual consent to meet the consent exception under the
Wiretap Act. The court provides the example that ``when someone
voluntarily participates in a telephone conversation knowing that the
call is being intercepted, this conduct supports a finding of implied
consent to the interception.''); Berry v. Funk, 146 F.3d 1003, 1011
(D.C. Cir. 1998) (The court found that ``[w]ithout actual notice,
consent can only be implied when the surrounding circumstances
convincingly show that the party knew about and consented to the
interception.'' In this case, the government tried to show implied
consent by arguing that any reasonable person would assume that an
operator stayed on the line if not told otherwise. In dismissing this
argument, the court found that ``[t]he key question in such an inquiry
obviously is whether parties are given sufficient notice.''); Deal v.
Spears, 980 F.2d 1153 (8th Cir. 1992) (holding that a warning about the
possibility, rather than actual notice, of monitoring did not
constitute sufficient notice for implied consent); Specht v. Netscape
Commc'ns Corp., 306 F.3d 17 (2d Cir. 2002) (evaluating claims outside
the Wiretap Act and considering the sufficiency of notice provided by
an inconspicuous software download license agreement); United States v.
Lanoue, 71 F.3d 966, 981 (1st Cir. 1995) (standing for the general
proposition that ``[d]eficient notice will almost always defeat a claim
of implied consent,'' this acknowledges that sufficient notice will
support a finding of implied consent).
---------------------------------------------------------------------------
B. State Wiretap Statutes
In addition to noting various all-party consent states, the CDT has
focused on California. First, the CDT Memorandum contains hedging
language regarding the application of various exceptions, such that if
such an exception were met, the issue of all-party consent would not be
reached as NebuAd's service would be exempted from liability under
alternative grounds.\17\ Additionally, the CDT notes the lack of
developed case law on the extension of the California wiretap statute
to Internet communications.\18\ The CDT points to the extraterritorial
application of the California wiretap statute involving telephone
communications--notably after recognizing that such statute may be
inapplicable to NebuAd's service--but then offers a countervailing
argument that California's consent requirement may be inapplicable to
behavioral advertising altogether.\19\ While focusing on the California
wiretap statute as a roadblock to NebuAd's service, the CDT Memorandum
itself recognizes that the statute may exempt NebuAd's service or may
be inapplicable to the industry altogether.
---------------------------------------------------------------------------
\17\ See supra notes 4-8 and accompanying text.
\18\ CDT Memorandum at 33.
\19\ Id.
---------------------------------------------------------------------------
______
Response to Written Question Submitted by Hon. David Vitter to
Clyde Wayne Crews, Jr.
Question. As our Committee continues to examine the issue of online
privacy, should we focus on the variations in different technologies
used to provide what appears to be essentially similar marketing
services? Or, should we instead focus on the use of the information
collected--by whatever method it is collected--to ensure that data is
used for legitimate marketing purposes, that privacy is protected, and
that we can go after those who misuse any data they collect?
Answer. Targeted marketing encompasses a diverse array of
relationships between users, marketers, and websites. While all types
of behavioral advertising techniques are often lumped together, there
is in fact a great deal of variety among online marketing business
models.
Even across search engines, there is no uniform method of dealing
with user data from search queries. Google, for example, to improve
search accuracy by recording every search query along with the user's
IP address and the time the search was conducted. Yet, Google does not
utilize this data to create user profiles. Another search provider, the
recently launched Cuil, does not retain any personal data. Microsoft
recently unveiled plans to deliver search results that take into
consideration each user's individual browsing habits, arguing that
Microsoft's Live Search engine will be better equipped to compete
against Google if it can analyze the intent of each user.
The misuse of sensitive information is an important concern for
policymakers evaluating online privacy issues. However, misuse can only
be defined on a case-by-case basis. This cannot be accomplished by
prescriptive legislation--especially in frontier sectors like the
Internet. The danger of stifling nascent markets is far greater than
any potential benefits of legislation which privacy mandates could
bring.
Controversial practices like the use of deep-packet inspection by
Internet Service Providers may offer commercial opportunities with
benefits to consumers. Lower broadband bills are just one possible
benefit from the delivery of personalized ads to subscribers. Of
course, for many users, privacy concerns trump potential benefits. In
some instances, firms may inaccurately assess consumer preferences,
resulting in mistakes like the NebuAd scuffle. Such mistakes, however
disturbing, will be resolved by market forces as competing firms
respond to consumer concerns by revising data collection practices as
needed.
Consider two users with different levels of concern regarding
personal privacy. One user lists his hobbies, friends, and demographic
details on a public social networking profile and does not expect that
data to remain private. On the other hand, another user whose Gmail
inbox contains sensitive personal correspondence assumes the data will
not be made public. In the current environment--where government
enforces voluntary privacy arrangements, but does not dictate them--
both users can be satisfied. Thanks to Gmail's robust privacy policy,
the concerned user can rest assured that her e-mails will be safe from
outside prying. And the user of the social networking site can enjoy
the services and content sustained by marketing income without
undesired rules preventing the use of data that is clearly in the
public sphere.
The technologies that drive online advertising are incredibly
complex, and new ways of analyzing data are constantly being developed.
Therefore, the fundamental question in the online privacy debate is how
we arrange for our data to be used once it has been transferred to a
third party. Calls for Congress to build walls around personal
information would preclude these arrangements, giving us too little
privacy in some cases and too much in others. A set of guidelines that
seems reasonable when applied to ISP deep-packet inspection might
eliminate other, more innocuous business models that rely on targeted
marketing.
______
Response to Written Questions Submitted by Hon. David Vitter to
Michael D. Hintze
Thank you again for your interest in the important privacy
implications of online advertising. Microsoft has a deep and
longstanding commitment to consumer privacy issues, and we welcomed the
opportunity to testify before the Senate Commerce Committee about the
concrete steps we are taking to protect consumers' privacy online. As
we indicated in our testimony, Microsoft believes that strong privacy
protections are not only compatible with bringing the benefits of
online advertising to consumers, advertisers and publishers, but are
essential to ensuring the success of this important business model.
This means that Microsoft, and all companies operating online, must
adopt meaningful privacy practices that build trust with consumers. We
believe our responses to your important follow-up questions demonstrate
that Microsoft takes this responsibility seriously.
Question 1. Does your company's business model already accommodate
the FTC's proposed principles for industry self-regulation? If so,
please explain how.
Answer. Yes, Microsoft's business model accommodates and even
exceeds the Federal Trade Commission's proposed principles for self
regulation. In our comments to the FTC, we urged the Commission to
focus on a broad array of online advertising activities (not simply
behavioral advertising) because all online advertising involves the
collection of data about computer users and may be contrary to
consumers' expectations.\1\ To this end, Microsoft specifically
advocated for a tiered approach to self regulation that is
appropriately tailored to account for the types of information being
collected and how that information will be used. Our proposal would
establish a baseline set of privacy protections applicable to all
online advertising activities and additional obligations for those
companies that engage in practices that raise additional privacy
concerns.
---------------------------------------------------------------------------
\1\ See http://www.ftc.gov/os/comments/behavioraladprinciples/
080411microsoft.pdf.
---------------------------------------------------------------------------
Microsoft's broad approach to self regulation is based on the
comprehensive privacy principles for online search and ad targeting we
announced in July 2007.\2\ These principles include commitments to user
notice, user controls, anonymization, security, and best practices.
Microsoft has embraced these privacy principles, and they will shape
the development of our new product offerings. We also have released a
set of privacy guidelines designed to help developers build meaningful
privacy protections into their software programs and online
services.\3\ The following paragraphs highlight the ways in which we
have implemented our own privacy principles into practice and, by doing
so, have also accommodated the FTC's proposed principles for industry
self-regulation.
---------------------------------------------------------------------------
\2\ Microsoft's Privacy Principles for Live Search and Online Ad
Targeting are available at http://www.microsoft.com/mscorp/twc/privacy/
default.mspx.
\3\ Microsoft's Privacy Guidelines for Developing Software Products
and Services are available at http://www.microsoft.com/privacy.
---------------------------------------------------------------------------
A. Transparency. Microsoft agrees with the FTC that transparency is
critical to enable consumers to make informed choices. To this end,
Microsoft's Online Privacy Statement is readily accessible from every
page of our websites, including the home page. It also is written in
clear language and offered in a ``layered'' format that provides
consumers with the most important information about our privacy
practices upfront, followed by additional layers of notice that provide
a more comprehensive examination of our general privacy practices.\4\
With respect to the delivery of advertisements online, the Microsoft
Online Privacy Notice Highlights clearly informs users about
Microsoft's online advertising practices, noting that Microsoft
``use[s] cookies and other technologies to keep track of your
interactions with our sites and services to offer a personalized
experience'' and that Microsoft's services ``may include the display of
personalized content and advertising.'' In addition, our full privacy
statement includes complete descriptions of the types of information
collected for online advertising and the ways in which such information
may be used. We believe our upfront and more detailed privacy
statements help ensure consumers are fully informed of our data
collection and usage practices.
---------------------------------------------------------------------------
\4\ Microsoft's Online Privacy Statement can be found at http://
go.microsoft.com/fwlink/?LinkId=74170.
---------------------------------------------------------------------------
B. Consumer Control. Microsoft also agrees with the FTC that the
collection of information about consumers to generate a profile of
their behavior upon which ads can be targeted raises heightened
concerns that warrant additional levels of user control. For this
reason, Microsoft has taken the following steps:
Microsoft was the first major online advertising provider to
announce it would give customers the opportunity to opt out of
receiving targeted advertising on all of the websites where
Microsoft provided advertising, including both Microsoft sites
and third-party partner websites.
Microsoft prominently provides information and links to our
opt-out mechanism in the top-layer of our privacy statement and
in our full privacy statement.
Microsoft allows users to tie their opt-out choice to their
Windows Live ID so their choice will be effective across
multiple computers without any additional effort on the user's
part.
Microsoft's opt-out method is more persistent than others--
for example, deleting cookies will not erase consumer's opt-out
selection; rather, their opt-out choice will be reset when they
sign in with their Windows Live ID.
We also recently announced three features of our new Internet
Explorer product that will improve consumer control. First, users may
choose to activate InPrivate browsing so their web surfing history,
temporary Internet files, and cookies are not recorded on their
computer after browsing. Second, users are given notice and a choice
about whether they want to block content coming from third parties that
may track and aggregate their online behavior by using the InPrivate
Blocking feature. Third, users have the choice to clear all or some of
their browsing history by using the enhanced Delete Browsing History
feature.
C. Security. Microsoft is committed to the FTC's principles around
data security. We have adopted strong data security practices,
implemented meaningful data protection and security plans, and
undertaken detailed third-party audits. We also have taken steps to
educate consumers about ways to protect themselves while online, and we
have worked closely with industry members and law enforcement around
the world to identify security threats, share best practices, and
improve our coordinated response to security issues.
D. Data Retention. Microsoft supports the FTC's principle that
entities that collect data through online advertising ``should retain
data only as long as is necessary to fulfill a legitimate business or
law enforcement need.'' As the Commission notes, there are often sound
and legitimate business reasons for retaining data collected from
users. These reasons include enhancing fraud detection efforts, helping
guard consumers against security threats, understanding website usage,
improving the content of online services, and tailoring features to
consumer demands.
Microsoft's policy around retaining search query data provides a
good example of the careful balance of interests that must be taken
into account when analyzing retention periods. Specifically, Microsoft
has committed to make search query data anonymous after 18 months by
permanently removing cookies, the entire IP address, and other
identifiers from search logs, unless the user has provided consent for
us to retain data for a longer period of time. Unlike other companies,
our anonymization method involves irreversibly removing the entire IP
address and other cross-session identifiers, such as cookies and other
machine identifiers, from search terms. Some companies remove only the
last few digits of a consumer's IP address, which means that an
individual search query may still be narrowed down to a small number of
computers on a network. We think that such partial methods do not fully
protect consumer privacy, so we have chosen an approach that renders
search terms truly and irreversibly anonymous.
E. Use of Personal Information for Online Advertising. Microsoft
agrees with the FTC that the merger of personally identifiable
information with other information collected about consumers through
behavioral advertising for the purposes of ad targeting presents
further privacy risks. This is because consumers are unlikely to expect
that a third party may combine such pieces of information and use it to
deliver ads (whether online or offline). For this reason, Microsoft has
developed its online ad targeting platform to select appropriate ads
based only on data that does not personally and directly identify
individual users, and we take steps to separate the data used for ad
targeting from any personally identifiable information before using it
to serve ads--a process we refer to as ``deidentification.'' \5\
Specifically, for users who have created Windows Live accounts, rather
than using the account ID as the basis for our ad systems, we use a
one-way cryptographic hash to create a new anonymized identifier. We
then use that identifier, along with the non-identifiable demographic
data, to serve ads online. Search query data and web surfing behavior
used for ad targeting is associated with this anonymized identifier
rather than an account identifier that could be used to personally and
directly identify a user.
---------------------------------------------------------------------------
\5\ Microsoft's ``de-identification'' white paper is available at
http://www.microsoft.com/privacy.
Question 2. Does your system accommodate for a consumer's choice
not to receive behavioral advertising, and in your systems, is that
request honored permanently? If so, please explain how.
Answer. Yes, Microsoft's system does accommodate for a consumer's
choice not to receive behavioral advertising. In July 2007, Microsoft
was the first major online advertising provider to announce it would
give customers the opportunity to opt out of receiving targeted
advertising on all of the websites where Microsoft provides
advertising, including both Microsoft sites and third-party partner
websites. This opt-out option became available in the Spring of 2008.
We prominently provide information and links to our opt-out mechanism
in the top-layer of our privacy statement and in our full privacy
statement.
Microsoft's opt-out choice is also unique from any other offered in
industry today because it is more persistent and applies across
multiple computers. As background, the industry-standard approach for
offering an opt-out choice is merely to place an ``opt-out'' cookie on
their machines. While this process generally works well, it does have
some inherent limitations. For example, opt-out cookies are computer-
specific--if a consumer switches computers, he or she will need to
specify any opt-out preferences again. Similarly, if cookies are
deleted, that user's opt-out choice is no longer in effect. To address
these limitations, the mechanism Microsoft offers gives consumers the
option to associate their opt-out choice to their Windows Live ID. This
means that even if they delete cookies on their machine, when they sign
back in their opt-out choice will persist. It also means that a single
choice can apply across multiple computers that they use. This will
help ensure that consumers' choices are respected without requiring
undue effort on their part.\6\
---------------------------------------------------------------------------
\6\ Microsoft's opt-out page is available at https://
choice.live.com/advertisementchoice/
Default.aspx.
---------------------------------------------------------------------------
Microsoft appreciates the opportunity to provide more information
about our privacy practices. We look forward to continuing to work with
you and all stakeholders to ensure consumers' privacy is protected
online.