[House Hearing, 110 Congress]
[From the U.S. Government Publishing Office]
ADDRESSING THE NATION'S CYBERSECURITY CHALLENGES: REDUCING
VULNERABILITIES REQUIRES STRATEGIC INVESTMENT AND IMMEDIATE ACTION
=======================================================================
HEARING
before the
SUBCOMMITTEE ON EMERGING
THREATS, CYBERSECURITY AND
SCIENCE AND TECHNOLOGY
of the
COMMITTEE ON HOMELAND SECURITY
HOUSE OF REPRESENTATIVES
ONE HUNDRED TENTH CONGRESS
FIRST SESSION
__________
APRIL 25, 2007
__________
Serial No. 110-30
__________
Printed for the use of the Committee on Homeland Security
[GRAPHIC] [TIFF OMITTED] CONGRESS.#13
Available via the World Wide Web: http://www.gpoaccess.gov/congress/
index.html
__________
U.S. GOVERNMENT PRINTING OFFICE
43-566 WASHINGTON : 2009
-----------------------------------------------------------------------
For sale by the Superintendent of Documents, U.S. Government Printing
Office Internet: bookstore.gov Phone: toll free (866) 512-1800
Fax: (202) 512-2250 Mail: Stop IDCC, Washington, DC 20402-0001
COMMITTEE ON HOMELAND SECURITY
BENNIE G. THOMPSON, Mississippi, Chairman
LORETTA SANCHEZ, California, PETER T. KING, New York
EDWARD J. MARKEY, Massachusetts LAMAR SMITH, Texas
NORMAN D. DICKS, Washington CHRISTOPHER SHAYS, Connecticut
JANE HARMAN, California MARK E. SOUDER, Indiana
PETER A. DeFAZIO, Oregon TOM DAVIS, Virginia
NITA M. LOWEY, New York DANIEL E. LUNGREN, California
ELEANOR HOLMES NORTON, District of MIKE ROGERS, Alabama
Columbia BOBBY JINDAL, Louisiana
ZOE LOFGREN, California DAVID G. REICHERT, Washington
SHEILA JACKSON LEE, Texas MICHAEL T. McCAUL, Texas
DONNA M. CHRISTENSEN, U.S. Virgin CHARLES W. DENT, Pennsylvania
Islands GINNY BROWN-WAITE, Florida
BOB ETHERIDGE, North Carolina MARSHA BLACKBURN, Tennessee
JAMES R. LANGEVIN, Rhode Island GUS M. BILIRAKIS, Florida
HENRY CUELLAR, Texas DAVID DAVIS, Tennessee
CHRISTOPHER P. CARNEY, Pennsylvania
YVETTE D. CLARKE, New York
AL GREEN, Texas
ED PERLMUTTER, Colorado
VACANCY
Jessica Herrera-Flanigan, Staff Director & General Counsel
Rosaline Cohen, Chief Counsel
Michael Twinchek, Chief Clerk
Robert O'Connor, Minority Staff Director
______
SUBCOMMITTEE ON EMERGING THREATS, CYBERSECURITY, AND SCIENCE AND
TECHNOLOGY
JAMES R. LANGEVIN, Rhode Island, Chairman
ZOE LOFGREN, California MICHAEL T. McCAUL, Texas
DONNA M. CHRISTENSEN, U.S. Virgin DANIEL E. LUNGREN, California
Islands GINNY BROWN-WAITE, Florida
BOB ETHERIDGE, North Carolina MARSHA BLACKBURN, Tennessee
AL GREEN, Texas PETER T. KING, New York (Ex
VACANCY Officio)
BENNIE G. THOMPSON, Mississippi (Ex
Officio)
Jacob Olcott, Director & Counsel
Dr. Chris Beck, Senior Advisor for Science & Technology
Carla Zamudio-Dolan, Clerk
Dr. Diane Berry, Minority Senior Professional Staff Member
(II)
C O N T E N T S
----------
Page
STATEMENTS
The Honorable James R. Langevin, a Representative in Congress
from the State of Rhode Island, and Chairman, Subcommittee on
Emerging Threats, Cybersecurity, and Science and Technology:
Oral Statement................................................. 1
Prepared Statement............................................. 3
The Honorable Michael T. McCaul, a Representative in Congress
from the State of Texas, and Ranking Member, Subcommittee on
Emerging Threats, Cybersecurity, and Science and Technology.... 4
The Honorable Bob Etheridge, a Representative in Congress from
the State of North Carolina.................................... 39
The Honorable Al Green, a Representative in Congress from the
State of Texas................................................. 37
Witnesses
Dr. Daniel E. Geer, Jr., Principal, Geer Risk Services, LLC:
Oral Statement................................................. 11
Prepared Statement............................................. 14
Dr. James Andrew Lewis, Director and Senior Fellow, Technology
and Public Policy Program, center for Strategic and
International Studies:
Oral Statement................................................. 6
Prepared Statement............................................. 8
Dr. Douglas Maughan, Program Manager, Cyber Security R&D,
Department of Homeland Security, Science and Technology
Directorate:
Oral Statement................................................. 23
Prepared Statement............................................. 25
Mr. O. Sami Saydjari, President, Professionals for Cyber Defense
Chief Executive Officer, Cyber Defense Agency, LLC:
Oral Statement................................................. 16
Prepared Statement............................................. 18
Appendixes
Appendix I: For the Record
The Honorable Bennie G. Thompson, a Representative in Congress
from the State of Mississippi, and Chairman, Committee on
Homeland Security, Opening Statement......................... 43
Appendix II: Selected Major Reports on Cyber Security Research
and Development................................................ 45
ADDRESSING THE NATION'S CYBERSECURITY CHALLENGES: REDUCING
VULNERABILITIES REQUIRES STRATEGIC INVESTMENT AND IMMEDIATE ACTION
----------
Wednesday, April 25, 2007
U.S. House of Representatives,
Committee on Homeland Security,
Subcommittee on Emerging Threats, Cybersecurity,
and Science and Technology,
Washington, DC.
The subcommittee met, pursuant to call, at 1:11 p.m., in
room 1539, Longworth House Office Building, Hon. James R.
Langevin [chairman of the subcommittee], presiding.
Present: Representatives Langevin, Etheridge, Green and
McCaul.
Mr. Langevin. The subcommittee will come to order.
The subcommittee is meeting today to receive testimony on
Addressing the Nation's Cybersecurity Challenges: Reducing
Vulnerabilities Requires Strategic Investment and Immediate
Action.
Good afternoon, and I want to welcome you to the
Subcommittee on Emerging Threats, Cybersecurity, and Science
and Technology hearing on a need to reduce vulnerabilities in
our national critical infrastructure through investment and
action. I would like to begin by thanking witnesses who appear
before us today, and I appreciate your testimony.
I think that last week was certainly an eye-opening
experience for many of us up here. We learned that our Federal
systems, in particular, and privately owned critical
infrastructure are all extremely vulnerable to hacking. These
vulnerabilities have significant and dangerous consequences.
We learned that the Federal Government has little
situational awareness of what is going on inside our systems.
We cannot be sure how much information has been lost from our
Federal systems, and we have no idea if hackers are still
inside our systems, and we learned that our laws are powerless
to stop intruders, even if compliance with FISMA does not make
our systems more secure--I should say even if best compliance
with FISMA doesn't make our systems more secure.
Now, this week, we are going to continue our conversation
from last week to hear about some promising initiatives that
are designed to reverse this trend of government failure.
I would like to take the opportunity to particularly thank
Dr. Maughan for his service to our country in this field. Dr.
Maughan is leading the cybersecurity research and development
effort at the Department of Homeland Security Science and
Technology Directorate. Under his leadership, DHS S&T has
funded research that has resulted in almost one dozen open
source and commercial products that provided capabilities such
as secure thumb drives, root kit detection, worm and
distributed denial of service detection, defenses against
phishing, network vulnerability assessment, software analysis
and security for process control systems.
His research and development funding is targeting the
critical problems that threaten the integrity, availability and
reliability of our networks. Clearly, he plays a vital role in
securing our natural cyberspace.
But despite the criticality of this mission and the success
of the program, I am troubled that this administration
continues its effort to do what Chairman Thompson calls
homeland security on the cheap.
In the last 7 years, more than 20 reports from such
entities as InfoSec, Research Council, the National Science
Foundation and the National Institute of Justice, the National
Security Telecommunications Advisory Committee, the National
Research Council and the President's Commission on Critical
Infrastructure Protection have all urged the government to do
more to drive, discover and deliver new solutions to address
cyber vulnerabilities. But look at what this administration has
done to cybersecurity and the research budget at the Department
of Homeland Security.
Though this program was slated to receive $22.7 million in
fiscal year 2007, the actual number I received from S&T showed
we only funded this program at $13 million. For fiscal year
2008, the President slashed the budget again, requesting $14.8
million. This is an $8 million cut from the previous year.
Just listen to some of the important programs that are
being cut or reduced in fiscal year 2007: The budget for the
DNSSEC program, which adds security to the main system, was
reduced $670,000. The budget for the secure protocols for
routing infrastructure was zeroed out from its original amount
of $2.4 million. The budget for the next generation
cybersecurity technologies program, which addresses a variety
of topic areas aimed at preventing, protecting against,
detecting, responding to and recovering from large-scale high-
impact cyber attacks, was reduced $1,625,000.
Now, I don't know who is responsible for these cuts, Under
Secretary Cohen or Secretary Chertoff or the White House, but
reducing this funding is a serious strategic error by this
administration.
Just to understand how little we are spending, for the sake
of comparison, the FBI estimated that, in 2004, that cyber
crime cost companies worldwide around $400 billion. In 2005,
the agency estimated that U.S. businesses lost $67 billion. Of
course, neither of these figures can measure the loss of
Federal information off our networks which one day may cost us
our technological advantage over other nations. And those
figures don't count the potential environmental losses if a
successful attack on our control systems were to be carried
out.
I am deeply troubled by the lack of foresight this
administration has demonstrated. These efforts are simply too
important to be cut.
The Homeland Security Committee is working to demonstrate
the importance of R&D funding in this administration. In our
recent authorization bill, we included a provision that would
increase the funding level for the DHS cybersecurity R&D
portfolio to $50 million. Democratic efforts over the last
several years have been endorsed by many notable cyber experts,
and I appreciate all of their input and their support.
The tools that will improve or revolutionize our security
will not just appear overnight. Investment today plants seeds
for the future. But it is incumbent upon the Federal Government
to take the leadership role in this effort.
Again, I want to thank our witnesses for appearing before
us today, and I look forward to hearing your testimony.
Prepared Opening Statement of the Honorable James R. Langevin, Chairman
Subcommittee on Emerging Threats, Cybersecurity, and Science and
Technology
Ladies and gentlemen, welcome to the Subcommittee on
Emerging Threats, Cybersecurity, Science and Technology hearing on the
need to reduce vulnerabilities in our national critical infrastructure
through investment and action.
I'd like to begin by thanking the witnesses who appear
before us today, and I appreciate your testimony.
I think last week was an eye opening experience for many
of us up here.
We learned that our federal systems and privately owned
critical infrastructure are all extremely vulnerable to hacking. These
vulnerabilities have significant and dangerous consequences.
We learned that the federal government has little
situational awareness of what is going on inside our systems. We cannot
be sure how much information has been lost from our federal systems,
and we have no idea if hackers are still inside our systems.
And we learned that our laws are powerless to stop
intruders--even the best compliance with FISMA does not make our
systems more secure.
This week, we're going to continue our conversation from
last week, and hear about some promising initiatives that are designed
to reverse this trend of government failure.
I'd like to take the opportunity to particularly thank Dr.
Maughan (``MAWN'') for his service to our country in this field.
Dr. Maughan is leading the cybersecurity research and
development effort at the Department of Homeland Security's Science and
Technology Directorate.
Under his leadership, DHS S&T has funded research has
resulted in almost one dozen open-source and commercial products that
provide capabilities such as:
secure thumb drives,
root kit detection,
worm and distributed denial of service detection,
defenses against phishing,
network vulnerability assessment,
software analysis, and
security for process control systems.
His research and development funding is targeting the
critical problems that threaten the integrity, availability, and
reliability of our networks. Clearly, he plays a vital role in securing
our national cyberspace.
But despite the criticality of this mission and the
success of the program, I am troubled that this Administration
continues its effort to do what Chairman Thompson calls ``Homeland
Security on the Cheap.''
In the last seven years, more than 20 reports from such
entities as the INFOSEC Research Council, the National Science
Foundation, the National Institute of Justice, the National Security
Telecommunications Advisory Committee, the National Research Council
and the President's Commission on Critical Infrastructure Protection
have all urged the government to do more to drive, discover and deliver
new solutions to address cyber vulnerabilities.
But look at what this Administration has done to
cybersecurity and the research budget at the Department of Homeland
Security.
Though this program was slated to receive $22.7 million
dollars in FY 2007, the actual numbers I've received from S&T show that
we are only funding this program at $13 million dollars.
For FY 2008, the President slashed the budget again,
requesting $14.8 million dollars. This is an $8 million cut from the
previous year.
Just listen to some of the important programs that are
being cut or reduced in FY 2007:
The budget for the DNSSEC program--which adds security
to the Domain Name System--was reduced $670,000 dollars.
The budget for the Secure Protocols for the Routing
Infrastructure was zeroed out from its original amount of $2.4
million dollars.
The budget for the Next Generation Cyber Security
Technologies program, which addresses a variety of topic areas
aimed at preventing, protecting against, detecting, responding
to, and recovering from large-scale, high-impact cyber attacks
was reduced $1.625 million dollars.
Now I don't know who is responsible for these cuts--Under
Secretary Cohen, or Secretary Chertoff, or the White House--but
reducing this funding is a serious strategic error by this
Administration.
Just to understand how little we're spending for the sake
of comparison, the FBI estimated in 2004 that cybercrime cost companies
worldwide around $400 billion dollars. In 2005, the agency estimated
that U.S. businesses lost $67 billion dollars.
Of course, neither of these figures can measure the loss
of federal information off of our networks, which may one day cost us
our technological advantage over other nations.
And those figures also don't count the potential
environmental losses if a successful attack on our control systems is
carried out.
I am deeply troubled by the lack of foresight that this
Administration has demonstrated. These efforts are simply too important
to be cut.
The Homeland Security Committee is working to demonstrate
the importance of R&D funding to this Administration.
In our recent authorization bill, we included a provision
that would increase the funding level for the DHS cybersecurity R&D
portfolio to $50 million dollars.
Democratic efforts over the last several years have been
endorsed by many notable cyber experts, and I appreciate all of this
support.
Ladies and gentlemen, the tools that will improve or
revolutionize our security will not just appear overnight. Investment
today plants seeds for the future, but it is incumbent upon the Federal
government to take the leadership role in this effort.
I thank the witnesses for appearing before us today and
look forward to their testimony.
Mr. Langevin. It is now my pleasure to recognize the
ranking member, my partner in this effort in the subcommittee,
the gentleman from Texas, Mr. McCaul, for purposes of an
opening statement.
Mr. McCaul. Thank you, Mr. Chairman.
I want to commend you again for holding this set of
hearings on cybersecurity, which is a very important issue
that, in my view, has been overlooked to a large extent since
September 11th. Last week, we heard from several government
agencies about their experiences with hackers breaking into
their networks. It is a serious problem, and it is happening
more often than we realize. As I have said before, I believe a
cyber attack could be at least if not more devastating to our
country than a weapon of mass destruction.
Unfortunately, right now, we are not doing what we need to
do to defend ourselves from this threat. Today, we focus on how
we respond to these attacks and how we develop the tools and
procedures to protect the information upon which our Nation
depends. Securing our networks may not get as much attention as
going to war, but it is just as important when we consider the
aspect of cyber warfare and the lack of our preparedness.
We have gathered some of the best minds here today in this
country to discuss how we as a country should respond to this
challenge of defending our information systems, and I look
forward to their testimony.
After our hearing last week, I met with a number of CEOs of
leading cybersecurity companies and heard their perspectives on
this complex issue; and it is clear that we must marshal our
resources and focus on this problem. We have not provided
information security the attention it deserves; and with the
help of experts such as those we have before us here today, I
believe we can improve the situation and provide the sense of
urgency to stimulate new progress in securing the Nation's
information systems.
I thank the Chair, and I look forward to the testimony.
Mr. Langevin. I thank the gentleman.
All the members as they arrive will be allowed to submit,
according to the committee rules, opening statements for the
record, and then we will begin to questions after the
testimony.
Again, I would like to turn to our panel right now. I want
to welcome our first panel of witnesses.
Our first witness, James Lewis, directs CSIS Technology and
Public Policy Program. He is a senior fellow. Before joining
CSIS, he was a career diplomat who worked on a range of
national security issues during his Federal service.
Our second witness, Dr. Daniel Geer, spent 10 years in
clinical and research medical computing, followed by 5 years
running MIT's Project Athena. Afterwards, he worked in the
research division of the then Digital Equipment Corporation and
then a series of entrepreneurial endeavors.
Our third witness is Mr. Sami Saydjari, who is the founder
and chief executive officer of Cyber Defense Agency, creators
of systematic defenses for high-value systems against
aggressive cyber attack. Before founding this cyber defense
agency, Mr. Saydjari was a senior staff scientist in SRI
International's Computer Science Laboratory.
Our fourth witness, Dr. Douglas Maughan, is the Cyber
Security Program Manager at the Department of Homeland Security
Science and Technology Directorate. Prior to his appointment at
DHS, Dr. Maughan was a program manager in the Advanced
Technology Office of the Defense Advanced Research Projects
Agency, or DARPA.
Without objection, the witness's full statements will be
inserted into the record; and I will now ask each witness to
summarize the testimony for 5 minutes, beginning with Dr.
Lewis.
Before we do that, though, I just wanted to remind everyone
of the committee rules that testimony is supposed to be
submitted 48 hours in advance. DHS didn't get their testimony
in to us until about 7:30 this morning. And I have said before
I understand DHS and other government departments need to get--
it is not solely on the witness's shoulders to get it in. I
know OMB has to clear the testimony. But this is happening
regularly from DHS. And I know Chairman Thompson is doing an
internal investigation right now to find out what the problem
is. We just can't do business like this if we don't have
testimony in a timely fashion.
Mr. Langevin. With that, I will turn it over to Dr. Lewis
for your opening statement. Thank you.
STATEMENT OF JAMES ANDREW LEWIS, DIRECTOR AND SENIOR FELLOW,
TECHNOLOGY AND PUBLIC POLICY PROGRAM, CENTER FOR STRATEGIC AND
INTERNATIONAL STUDIES
Mr. Lewis. Thank you, Mr. Chairman. I would like to thank
the committee for this opportunity to testify.
You heard last week about the problems at various agencies,
and I think that testimony highlighted that securing networks
in the United States for cyber attack is one of the greatest
challenges we face.
Cyber security can seem intractable. It is a problem that
in the past attracted exaggeration, and this combination of
intractable and exaggeration can sometimes create indifference.
One way to overcome this indifference is to put cybersecurity
in the right context.
Our networks face two sets of risks. The first involves
espionage. The second involves the potential interruption of
services, particularly Federal services.
The most important for me is espionage cybersecurity, is
primarily a spy story. Cyber espionage poses the greatest
threat to the U.S. Hacking into computer networks, which are
vulnerable and likely to remain so for years, provides new low-
cost and low-risk opportunities for foreign intelligence
agencies. U.S. networks are very vulnerable. Several nations
have exploited these vulnerabilities to gain valuable
information. These efforts and our inadequate response have
damaged national security.
Unlike cyber espionage, the threat of disruption of
services remains hypothetical. I would not take too much
comfort from this, Mr. Chairman. Because if an opponent can
hack into a network to steal information, they can hack into
and plant malicious software that could be triggered during a
crisis. We should assume in the event of a conflict our
opponents will seek to disrupt our networks and data.
I would like to point out that, although we have a long
litany of threats, the question as to whether the U.S. was
better off before it depended so heavily on computer networks
can be answered in the negative. The benefits from the greater
use of networks and computers outweigh the damage from poor
cyber security. However the porousness of our Federal networks
reduces those benefits, and greater attention cybersecurity
would improve both national security and economic performance
and close off an avenue of asymmetric opportunity for our
opponents.
While the U.S. is better off than it was 10 years ago, the
improvement has been unequally distributed among agencies and
companies. Some are secure; some are not. There have been
serious efforts in the national security community to make
networks more secure, and our most sensitive military and
intelligence functions are probably secure. Some crucial civil
networks are also more secure than they were.
Some efforts to improve cybersecurity have not had the
benefits we expected. It is possible to hack into a computer
running software that has met the common criteria, that has the
common criteria certification, on a network that has met the
requirements of ISO 19779, the standard for cybersecurity, and
at an agency that has gotten good marks on FISMA. In other
words, you can meet all the formal requirements and still be
vulnerable.
How do we change this? There is no silver bullet. There is
no single program that will improve security. The Federal
Government, for example, is a complex enterprise, with
thousands of networks and hundreds of thousands of computers.
No single agency controls this network; and while some Federal
networks are among the most secure in the world, others are
routinely penetrated. Some use advanced technologies, others
are legacy systems dating back years and which, for all
practical purposes, cannot be secured.
The core of the problem is organizational. The Department
of Homeland Security, the Federal CIO Council, and the Office
of Management and Budget all play a role in securing Federal
networks. But cybersecurity remains a low priority at many
agencies.
Along with a better organization for cybersecurity, the
U.S. needs a better strategy. We did have a national
cybersecurity strategy in 2003, but it is outdated. A new
strategy would have to be more comprehensive, and I would like
to detail some of the things I think that strategy should
include.
First, we would benefit from streamlining government
processes. There are too many groups and committees, and too
few of them have any real authority.
Second, the U.S. can do more to improve agency practices
for network security. Cybersecurity is still a third-tier
priority at many agencies. If gangs of foreigners broke into
the State or Commerce Departments and carried off dozens of
file cabinets, there would be a huge outcry. When the same
thing happens in cyberspace, we shrug it off. Agencies need to
be held accountable for following best practices in network
security.
Third, better identity management would improve
cybersecurity security. As long as it is easy to impersonate
someone on the Internet, networks will never be secure. HSBD 12
and Real ID can offer some benefits.
Fourth, the government should address software assurance.
We recently did a study at CSIS that looked at how companies
write software. While most of them do a pretty good job and all
of them have some very useful practices, the practices aren't
evenly applied; and if the government could find a way to
spread these best practices to make software more secure, it
would have a real benefit.
Finally, the U.S. can take steps to keep itself at the
forefront of technology. This goes beyond funding cybersecurity
research. While we spend more on R&D than other countries, it
may not be enough to maintain our lead. These steps--better
organization, better practices for coding, better identity
management, attention to continuity of government and renewed
support for technological leadership--can make networks more
secure.
Congressional oversight is critical with this. Without
Congress to press senior leadership at Federal agencies, we
will wait much longer for progress than would otherwise be the
case.
It has been 12 years since the U.S. became concerned with
vulnerabilities in computer networks. There has been some
improvement, but not enough. We have an opportunity to change
this in the next few years.
I thank the committee for the opportunity to testify. Thank
you for entering my comments into the record, and I will be
happy to take your questions.
Mr. Langevin. Thank you, Dr. Lewis.
[The statement of Mr. Lewis follows:]
Prepared Statement of Dr. James A. Lewis
I would like to thank the Committee for the opportunity to testify
on the cybersecurity challenge the United faces. Cybersecurity is one
of those problems that seem to be intractable. It is also a problem
that, in the past, seemed to attract exaggeration and hyperbole. The
combination is not ideal for creating effective policies, in part
because the blend of intractability and exaggeration can create
indifference.
One way to overcome this indifference is to put cyber security in
the right context. The context is not an `electronic Pearl Harbor' but
the risk of loss of valuable information and the disruption of data and
services. For Federal networks, the context for cybersecurity involves
espionage and potential interruptions in the delivery of Federal
services.
The security of Federal networks has serious implications for
homeland security as Federal network security affects both continuity
of government and the operations of critical infrastructure. This alone
justifies extra attention to government networks. In addition, measures
that improve the security of Federal networks will also benefit private
sector networks. My own view is that the security of Federal networks
is the most serious cybersecurity challenge we face, more serious than
the risks to critical infrastructure or from cybercrime.
The most important of these challenges come from espionage.
Cybersecurity is primarily a spy story. Cyber-espionage poses the
greatest current threat to the United States. Hacking is the extension
of signals intelligence into new and untrammeled areas. Foreign
intelligence agencies must weep with joy when they contemplate U.S.
government networks. We have thoughtfully placed sensitive information
on these networks and then failed to secure them adequately. This is
not a hypothetical problem. The last twenty years have seen an
unparalleled looting of U.S. government's databases.
The reliance upon information technology has changed the nature of
espionage. Information is more valuable. Nations will use the
traditional means of espionage (infiltration and recruitment) to obtain
access to information, but information technologies have created a
gigantic new opportunity. Hacking into computer networks (which are
vulnerable and likely to remain so for years) provides new, low cost
and low risk opportunities for espionage. Eight or nine countries have
the advanced technical skills needed for these operations and smaller
countries could hire hackers from the criminal world--we know of at
least one instance where this has occurred.
Conflict in cyberspace is clandestine, so it can be difficult to
assess our opponents' intentions and capabilities. It is easier to
assess the vulnerability of U.S. systems and the consequences of an
information attack. U.S. networks are very vulnerable. Even highly
sensitive networks used for command and control or intelligence are not
invulnerable. From an intelligence perspective, several nations, have
exploited the vulnerabilities of U.S. government networks to gain
valuable information. These foreign intelligence efforts and the
inadequate U.S. response have damaged national security.
You heard last week about some of the problems that some agencies
face. Their testimony highlights that securing Federal networks from
cyber attack is one of the greatest challenges facing the United
States, and that the scope of the challenge and the threat to national
security are difficult to appreciate fully. Several incidents that
occurred in the past few months help to illustrate the scale of the
problem. In December and January 2006, for example, the Naval War
College, the National Defense University, and other DOD facilities had
to take computer networks offline after a foreign entity infected them
with spyware. Before the last shuttle launch, NASA had to block e-mail
attachments to avoid outsider attempts to gain access before a Shuttle
launch. And as you heard last week, the Department of Commerce had to
all of the computers at the Bureau of Industry and Security offline
after they were hacked and infected with spyware.
In contrast to espionage, the threat of the disruption of services
remains hypothetical. Cyber-espionage is a routine occurrence, but
there have been no disruption of services. We should not take much
comfort from this, however. If an opponent can hack in to Federal
networks to steal information, they are likely to also be able to hack
in to implant malicious software that could be triggered in a crisis to
disrupt services or to scramble data. It is safe to assume that many of
our potential opponents are planning informational attacks to disrupt
U.S. government services and databases.
It is easy to overstate the effect of this disruption, but a
cyberattack that increases uncertainty in the mind of an opponent
degrades that opponent's effectiveness. This is a classic intelligence
strategy, and cyber attacks on information systems provide new and
expanded means to execute it. Denial and deception can make opponents
certain that they know what is happening when, in fact, what they
believe is wrong, or it can make them unsure that they know what is
happening. Finding ways to inject false information into the planning
and decision processes of an opponent, or manipulating information that
is already in that system to make it untrustworthy, can provide
military advantage. In the event of a conflict, our opponents will
pursue an informational strategy that seeks to expand uncertainty and
confusion and this will likely involve efforts to disrupt Federal
networks.
This litany of threats and risks might lead some to ask if the U.S.
was better off before it depended so heavily on computer networks. The
answer to that question is no. The benefits to the U.S. that come from
the greater use of networks and computers outweigh the damage from poor
cybersecurity. It is better to have networks than to be without them,
and the use of computer networks provides the U.S an advantage in its
economy and its military operations. However, the porousness of our
Federal networks erodes those benefits. Greater attention to
cybersecurity would increase the benefits our nation gains from
networks and close off an avenue of asymmetric advantage to our
opponents.
There have been serious efforts in the national security community
to make their networks more secure. Our most sensitive military and
intelligence functions are probably secure. Some civil crucial networks
are more secure--much attention has been paid to Fedwire, the Federal
Reserves electronic funds transfer system, for example. But, as you
heard last week, many agency networks remain poorly secured, and it is
safe to say that reams of diplomatic, scientific, administrative and
defense industrial information at the various agencies have not been
adequately secured. In looking at the security of Federal networks, it
is fair to say that while the U.S. is better off than it was five years
ago or ten years ago, the improvement has been unevenly distributed
among agencies. Some are secure, most are not.
Additionally, some efforts to improve cybersecurity have not had
the benefits we expected. It is quite possible for our opponents to
hack a computer running software that has Common Criteria
certification, on a network that has met the requirements of ISO 19779,
at an agency that has gotten good marks on FISMA. In other words, you
can meet all the formal requirements and still be vulnerable.
Network security is also a dynamic situation, dynamic in the sense
that attacks are continuous and continuously changing. We should
applaud those agencies that have, after some months, discovered their
networks have been hacked and have taken steps to undo that hack, but
our next question should be, ``and now what are you doing.'' Attacks on
Federal networks are continuous, and fixing one problem does not mean
that we have checked the box and can turn our attention elsewhere.
How doe we change this situation? There is no silver bullet, no
single program or effort that will remedy this problem. Increased
funding will not improve security. The Federal Government is a complex
enterprise, with thousands of networks and hundreds of thousands of
computers. No single agency has control of this collection of networks.
Some Federal networks are among the most secure in the world, although
even these are not immune from attack. Others are routinely penetrated.
Some systems use the most advanced technologies. Others are legacy
systems, running programs that may date back many years and which, for
all practical purposes, cannot be secured.
Making networks more secure is a large and complex problem. The
core of the problem is organizational. Although it has been more than a
decade since the Marsh report on the risks posed by cyber attack to
critical infrastructure, and although there has been progress, the
Federal Government is still disorganized when it comes to cyber
security. The Department of Homeland Security, the Federal CIO Council,
and the Office of Management and Budget all play a role in securing
Federal networks. But cybersecurity remains a low priority and an
afterthought for many agencies, and the Federal response to
cybersecurity remains largely ad hoc and dispersed.
Along with better organization, the U.S. also needs a better
strategy. There is, of course, a National Cyber Strategy from 2003, but
that strategy is now outdated. It shifted too much of the burden for
security to the private sector and did not resolve key issues regarding
responsibility within the government. A new, comprehensive cyber
security strategy for the Federal Government would need to include a
number of complementary measures to reduce vulnerabilities. The
following paragraphs provide a brief outline of some of the major
elements of this approach.
Rationalizing and streamlining governmental processes for improving
cybersecurity is essential. There are too many interagency groups and
committees working on the same problem, often with the same people, and
few of them have the authority to make any real progress. The U.S. does
not need a new White House cyber czar, but it does need to do more to
direct and coordinate efforts by the various agencies. The recent
creation of a cybersecurity Policy Coordinating Committee at the
National Security Council is an important first step.
Second, the U.S. can do more in the area of improving agency
practices when it comes to networks security. Cybersecurity is still a
third tier priority at many agencies. If gangs of foreigners broke into
the State or Commerce Departments and carried off dozens of file
cabinets, there would be a crisis. When the same thing happens in
cyberspace, we shrug it off as another of those annoying computer
glitches we must live with. Agencies need to be held accountable for
breaches. Our current approach is to treat losses of information
through inadequate security as something that is separate from the
performance of senior officials.
The separation between the national security agencies and civilian
agencies needs to be reduced. The national security agencies do better
at security, but there is no good mechanism for sharing their expertise
and experience with the civilian agencies. Developing better ways to
coordinate network security efforts between agencies and to identify,
share and enforce best practices for Federal network security across
agencies would reduce risk and damage.
Better identity management would also help improve cybersecurity.
As long as it is easy to impersonate someone else on the internet,
networks will never be secure. In this, initiatives like HSPD 12 and
the Real ID Act offer the possibility to reduce risk. HSPD-12 mandated
strong identity procedures and credential for the Federal Government
and its contractors HSPD-12, along with Real ID, lay the foundation for
robust authentication of identity. Much remains to be done, but the
U.S. has begun to adjust how it manages identities to fit digital
technologies and this will improve security.
Continued attention to continuity of operations and continuity of
government can mitigate the risk of disruption of Federal services. As
part of a Federal cybersecurity strategy, this would entail measures to
keep networks operating at some minimal level and to provide continued
access to data. This is an area where there has also been some
progress.
One new area the government can begin to address is how to improve
software assurance. This means creating processes for transparency,
evaluation and coordination in the production of more secure software
for government use. In considering this, let me refer to an episode
from American history, when the U.S. faced a similar problem and what
it did about it. This story has an unlikely hero--Herbert Hoover.
Hoover may have been a terrible or unlucky President, but he was a
great Secretary of Commerce. One of the things he did in the 1920's as
Secretary of Commerce was call a number of leading companies from
different sectors - automobiles, electrical equipment and so on, to the
Commerce Department and say that they had to come up with a means to
improve quality and interoperability in their products. This was the
start of the industry-led standards process.
We need something similar to happen for security and software
production. There are existing standards bodies for software. These
standards are aimed at products--how they perform and how they
interoperate. The U.S. does not need to duplicate them. What we need is
a new means for understanding how to produce software in ways that can
assure security.
CSIS recently did a study that looked at how some of the larger IT
companies write software. We found considerable attention to security
among the companies, and that each company had a set of `best
practices' for software assurance that make their products more secure.
We also found that each company's best practices were somewhat
different, and that these practices were sometimes unevenly applied.
Finding a way to extend commercial best practices for assurance
would benefit both Federal networks and the private sector. The
procedures companies use as part of their software production process
internal reviews and testing for performance and security, external
testing and red-teaming, and the use of software review tools (some
commercial, some proprietary and developed by the software company
itself) to find vulnerabilities or errors. These practices offer the
building blocks for an approach that could reduce vulnerabilities.
The key to these new processes should be to build upon what is
already done within the private sector when it comes to software.
Software producers realize the importance their customers place on
assurance and security and have adjusted their internal procedures to
meet this market demand. While there is much commonality and overlap in
what companies do, each company approaches the issues of assurance and
security somewhat differently. From these differences, we can extract
best practices and requirements that will address, as part of a larger
solution set, the risks posed by foreign involvement in software
production.
Please note that I am saying best practices, not standards. An
attempt to have the government mandate standards for software
production and then enforce them would damage the American economy
without producing any benefit for security. So new regulations, new
government standards are not the solution. However, the government
could encourage industry to use best practices for making secure
software by linking practices to its acquisitions policies. If the
Federal Government gave preference in its acquisitions to software that
was developed with trustworthy processes, it would provide an incentive
that would benefit both the Federal and the commercial markets.
Companies are making serious efforts to improve software assurance,
but the government needs to be able to understand and guide those
efforts. Traditional approaches to governance--command and control or
heavy regulation--would increase assurance at an unacceptable cost.
Software assurance may be the effort that promises the greatest returns
to cybersecurity. The U.S. needs news ways to let the government and
the private sector work together to develop some generalized set of
best practices for software production, and the Departments of Defense
and Homeland Security are involved in some interesting work in this
area.
Finally, the U.S. can take steps to keep itself at the forefront of
technology. This goes beyond simply funding more cyber-security
research. Overall, the U.S. invests more than other nations in
research, but this investment may not be enough, in an era of increased
international competition, to preserve leadership. Federal investment
in the research that undergirds technological innovation offers
tremendous returns for both the economy and for security. Innovation
makes life more difficult for opponents. Measures that improve the
climate for innovation in the U.S. also help build a skilled domestic
workforce.
These steps--better Federal organization, best practices for coding
combined with acquisitions, better identity management, attention to
continuity of government and renewed support for technological
leadership--can form a coherent strategy for improving the security of
Federal networks and cybersecurity in general. Being able to articulate
a strategy is important, but implementation will always be a challenge.
In this, Congressional oversight is critical to this. Without Congress
to press senior leadership at Federal agencies to do better, progress
will take much longer than would otherwise be the case.
It has been more than twelve years since the U.S. became concerned
with the vulnerabilities created by its use of computer networks. There
has been some improvement in that time, but not enough. We have an
opportunity in the next few years to change this with improved Federal
organization and better strategies. Our goal should not be perfect
security, but to gain more advantage than our opponents from the use of
information technology.
I thank the committee again for the opportunity to testify. I ask
that my entire statement be entered into the record, and I will be
happy to take your questions.
Mr. Langevin. Dr. Geer.
STATEMENT OF DANIEL E. GEER, JR., PRINCIPAL, GEER RISK
SERVICES, LLC
Mr. Geer. Thank you.
I don't do this every day, so I am just going to start with
what I know of as the four verities of government, which is
most exciting ideas are not important, most important ideas are
not exciting, not every problem has a good solution, and every
solution has a side effect. And that is amazingly true in the
field that I work in, cybersecurity. Every bit of that is true.
I am going to try to give you five priorities from my point
of view.
The first is, we need a system of security metrics, metrics
that actually work. One of the great scientists of all time,
Lord Calvin, said, and I have to read this:
When you can measure what you are talking about, and
express it in numbers, you know something about it; but when
you cannot measure it, when you cannot express it in numbers,
your knowledge is a meager and unsatisfactory sort; it may be
the beginning of knowledge, but you have scarcely, in your
thoughts, advanced to the stage of science.
As we stand here today, we do have some metrics. Most of
them are imperfect--all of them are imperfect. A few are good
enough for decision making.
In late 2003, the NSF held a sequestered invitation-only
workshop to determine the 10-year ``grand challenges'' in
cybersecurity. One of those four grand challenges that we came
up with one speaks directly to this: Within a decade, we must
have a body of quantitative information risk management as
sophisticated as quantitative financial risk management. That
item actually was mine, and it was my pleasure to present it to
House Science.
Good metrics aren't cooked in the kitchen. They don't
appear on demand. Like statistics, they can mislead. The
purpose of risk management is to improve the future, not to
explain the past. Security metrics are the servants of risk
management, and risk management is about making decisions under
uncertainty. Therefore, the metrics I am talking about, the
only ones we are interested in, are those that support decision
making about risk for the purpose of managing that risk.
I would recommend that some sort of clearinghouse review of
what we know how to measure and in particular how good what we
know how to measure is at predicting the future would be a good
thing to do right away.
Second priority. The demand for security expertise
outstrips the supply.
Information security is, in my view, the hardest technical
field on the planet. Nothing is stable, surprise is constant,
and defenders are at a permanent structural disadvantage
compared to the attack side. There is no fixing that.
But because the demand for expertise so outstrips the
supply, the fraction of practitioners who are charlatans is
rising. Because the demands of expertise are so difficult, the
training deficit is critical. We don't have the time to create
all the skills that are required. We have to steal them from
other fields.
The reason cybersecurity is not worse than it might
otherwise be is because a substantial majority of those who are
currently practicing were trained in other fields and,
therefore, they bring the expertise of those other fields to
this one. We are lucky that that is true. Civil engineers,
public health people, actuaries, aircraft designers, lawyers,
you name it, all of them can contribute something.
We do not have the facility to train people from scratch at
the rate at which we need it; and so anything you can do to
encourage people to come into this field who are themselves
smart, analytic, willing to operate under a high degree of
uncertainty and convinced that this is worth doing, anything
you can help with that, please do.
Third priority. What you can't see is more important than
what you can.
Perhaps you got a taste of it last week. I was not aware of
that hearing. I don't follow this kind of thing. Let me be
clear, the opposition is professional. It is not joyriders. It
is not braggarts. It used to be, but it isn't now. Because of
the sheer complexity of modern networks, there is any number of
places for people of ill-will or for computer software of ill-
will to hide. And that is not getting better, and it won't get
better.
The complexity for the most part is because product
manufacturers are under competitive pressure to keep inserting
new features into their products. This is not going to go away,
and it is not something I would suggest that you attack. Were
there no attackers, the way in which software is built would be
a miracle of efficiency. The fact that there are attackers, the
fact there are sentient opponents, the fact that this is not
evolution but intelligent design of a nasty sort, that is what
we have to work on.
Complex systems tend to fail in complex manners. It is very
hard to figure that out in advance. It is exceptionally hard.
That is why I say it is probably the hardest field there is.
In particular, I think what you need to do is to do
something that I don't like the sound of but I will say.
Ignorance of the law is no defense on my part. My swimming pool
is an attractive nuisance, whether I like it or not. I don't
think we can go much farther and say that I didn't know it had
a flaw is any kind of defense. And software licenses, to the
last one of them, have that built into them, and it has to be
addressed.
The fourth one is we have to have some sort of information
sharing. You all know about all of this. I am not going to
belabor it. The model I would recommend to you is the Centers
for Disease Control. They only have three things that matter:
the mandatory reporting of communicable disease, the skill to
separate statistical anomalies from true hot spots, and an away
team to handle things like an outbreak of ebola. Beyond that,
nothing matters.
I would suggest that something like that needs to be done
here. No general counsel acting rationally will ever share
attack data. There is nothing but downside risk from where they
are.
So if I can give you a research grade problem to work on,
the research grade problem is this: Find some way to do
technical de-identification of attack data so that general
counsel's rational fear of sharing that data can be put aside
under a technical guarantee. They do not and they will not
believe your procedural guarantees. We have got to have a
technical guarantee. This is a research grade problem that
needs to be done.
The fifth one and last one is perhaps the hardest of all,
and that is accountability rather than access control. Access
control is who you are, authentication, what you are allowed to
do given who you are, authorization. It doesn't scale. And if
we try to make it scale--that is not to say everybody does it
well as it is, but if we try to make that scale, the rate at
which data and facilities and knobs to adjust are increasing is
out of our ability to add to that full-blown access control
going forward.
We have to do something else. This is a free country. I
didn't have to ask anyone's permission to be here, to get on
the bus or what have you. But if I sufficiently badly screw up,
then I will have to pay for it. We are in the physical world
committed now to surveillance, whether we like it or not. You
can't live your life without metal detectors and cameras. We
are going have to do that in this world.
And if I may say so, please make sure that the surveillance
we have to do is directed at data and computers and not at
people. It is a choice we have to make, and it is an ugly
choice.
I will just say the five things again and be quiet.
We need a system of security metrics, and it is a research
grade problem.
The demand for security expertise outstrips the supply, and
it is both a training and a recruitment problem.
What you can't see is more important than what you can, and
you can never mistake the absence of evidence for the evidence
of absence.
Information sharing that matters does not happen and cannot
happen until we have technical guarantees, rather than
procedural ones.
And accountability is an idea whose time come, but--to
steal Leon Uris' phrase--it has a terrible beauty.
Thank you.
Mr. Langevin. Thank you, Dr. Geer.
[The statement of Mr. Geer follows:]
Prepared Statement of Daniel E. Geer
Introduction
The Nation's cybersecurity challenges are profound and not easily
addressed. Perfection is not possible; rather this is entirely a matter
of risk management, not risk avoidance. Easy to say.Hard, though not
impossible, to do. Starting yesterday would be good. Money alone will
not solve anything. Policy alone will not solve anything. Fixing what
isn't broken will waste money capital and policy capital; fixing what
is broken will require both. Wishful thinking, whether explicit or
implicit, intentional or delusional, will allow the problem to get
bigger.
In the testimony which follows, I make no attempt to argue from
first principles or to provide every supporting footnote that would be
required to prove the assertions made; I don't think you want it and
the page limit prevents it. I do, however, have all the proof that can
be had, and stake my professional reputation on what is said here. I
trust that you have invited me because you are aware of that reputation
and my bona fides in these matters. The material is brief in the hope
that brevity increases the likelihood it will be read. This is not your
last chance to get my attention; I hope it is not my last chance to get
yours.
Priority number one: A system of security metrics.
``You cannot manage what you cannot measure'' is a cliche, but,
happily,one of the great scientists of all time, William Thompson, Lord
Kelvin, put it as well as it can be put:
When you can measure what you are speaking about, and express
it in numbers, you know something about it; but when you cannot
measure it, when you cannot express it in numbers, your
knowledge is a meagre and unsatisfactory kind; it may be the
beginning of knowledge, but you have scarcely,in your thoughts,
advanced to the stage of science.
As we stand here today,we have some security metrics. None of them
are perfected though many are good enough for decision making if, and
only if, they are collected by persons whose aim is truth rather than
positioning. In late 2003, the Computing Research Association and the
National Science Foundation held an invitation-only workshop to
determine the ten-year ``grand challenges'' for NSF investment in
cybersecurity. Of the four grand challenges settled upon, one speaks
directly to this: Within a decade, we must have a body of quantitative
information risk management as sophisticated as the then existing body
of financial risk management. That item was mine, and I had the honor
of presenting it to this body immediately after the conclusion of the
workshop.
Good metrics are not cooked in the kitchen. They are not created
simply because the Congress demands them. Like statistics, they can
mislead. In your line of work, you doubtless know this better than I
and I know it well. The purpose of risk management is to improve the
future, not to explain the past. Security metrics are the servants of
risk management, and risk management is about making decisions under
uncertainty.Therefore, the only security metrics we are interested in
are those that support decision making about risk for the purpose of
managing that risk. I urge the Congress to put explaining the past,
particularly for the purpose of assigning blame, behind itself.
Demanding report cards, legislating under the influence of adrenaline,
imagining that cybersecurity is an end rather than merely a means--all
these and more inevitably prolong a world in which we are procedurally
correct but factually stupid. A clearinghouse review of what we know
how to measure and how good what we know is at predicting the future
would be a good start as we do not even know what it is that we do not
know.
Priority number two: The demand for security expertise outstrips the
supply.
Information security is perhaps the hardest technical field on the
planet. Nothing is stable, surprise is constant, and all defenders work
at a permanent, structural disadvantage compared to the attackers.
Because the demands for expertise so outstrip the supply,the fraction
of all practitioners who are charlatans is rising. Because the demands
of expertise are so difficult, the training deficit is critical. We do
not have the time to create, as if from scratch, all the skills
required. We must steal them from other fields where parallel
challenges exist. The reason cybersecurity is not worse is that a
substantial majority of top security practitioners bring other skills
into the field; in my own case, I am a biostatistician by training.
Civil engineers, public health practitioners, actuaries, aircraft
designers, lawyers, and on and on--they all have expertise we can use,
and until we have a training regime sufficient to supply the unmet
demand for security expertise we should be both grateful for the
renaissance quality of the information security field and we should
mine those other disciplines for everything we can steal. If you can
help bring people into the field, especially from conversion, then
please do so. In the meantime, do not believe all that you hear from
so-called experts. Santayana had it right when he said that
``Scepticism is the chastity of the intellect; it is shameful to give
it up too soon, or to the first comer.''
Priority number three: What you cannot see is more important than what
you can.
The opposition is professional. It is no longer joyriders or
braggarts. Because of the sheer complexity of modern, distributed,
interdigitated, networked computer systems, the number of hiding places
for unwanted software and unwanted visitors is very large. The
complexity,for the most part, comes from competitive pressure to add
feature-richness to products; there is no market-leading product where
one or a small group of people knows it in its entirety,and components
from any pervasive system tend to be used and re-used in ways that even
their designers did not anticipate. Were there no attackers, this would
be a miracle of efficiency and goodness. But unlike any other
industrial product, information systems are at risk not from accident,
not from cosmic radiation, and not from clumsy operation but from
sentient opponents. The risk is not, as some would blithely say,
``evolving'' if by evolving the speaker means to invoke the course of
Nature. The risk is due to intelligent design, and there is nothing
random about it.
Because complex systems fail complexly, it is not possible to
anticipate all the failure modes of large and therefore complex
information systems. This complexity provides both opportunity and
hiding places for attackers. Damping out complexity is not something
that even the Congress can take on, but security failures come from it
as surely as dawn comes from the east. Given that most software license
agreements are an outrage, it is high time that security failures in
software systems be deemed per se offenses. Just as my ignorance of the
law is no defense and my swimming pool is an attractive nuisance
whether I like it or not, ignorance of installed vulnerabilities can no
longer be a defense for any party.
Priority number four: Information sharing that matters.
On the Internet every sociopath is your next door neighbor; you can
never retreat to a safe neighborhood. Your ability to defend depends on
your ability to know what the current threat profile is, both generally
to all and specifically to yourself. For any given attack, you have
zero ability to know whether you are a target of choice or a target of
opportunity unless you share attack data with others.
Our Centers for Disease Control lead the world, full stop. There
are only three things that make this so: (1) Mandatory reporting of
communicable disease, (2) Longitudinal analysis and the skill to
separate statistical anomalies from genuine harbingers of important
change, and (3) Away teams to handle outbreaks of, say, Ebola. All the
rest is details. Of the three, the one that matters most is the
mandatory reporting of communicable disease, and explicitly on the
grounds that individual medical privacy must yield when the public risk
is above threshold.
No General Counsel will share information risk data willingly, and
no Chief Information Security Officer outranks his/her GC. Shared
information does always carry some acute chance that it contains a
previously unknown embarrassment, while any benefit from sharing is
diffuse and delayed. Any person is risk averse when they don't know
what risk they are taking and more so when the risk is involuntary; the
GC is rational to not share data, in other words. The Congress should
be wary of legislating irrationality, as always.
To get information shared the need is for a technical guarantee of
harmlessness rather than a procedural guarantee. This is, in other
words, a straight-up research question: How to provide technical de-
identification of useful cybersecurity data so that that data can be
shared with low or no risk to its source. Such technical protection
should be open-sourced so that its strength can be independently
evaluated a priori rather than the ``trust us'' nature of a procedural
guarantee. Fund this research.
Priority number five: Accountability,not access control.
Information is the coin of the economic realm, and information that
is used is information that moves about. Winners have the most
information in play; losers have too much. Security technology is the
fine line between the most information in play and too much information
in play. The conventional answer to protecting information is to in
some way limit who can do what and to which. Authentication (who you
are) and Authorization (what you can do, given who you are) represent
the conventional approach, sometimes jointly called Access Control. The
problem is, these technologies do not scale and if you try to have ever
finer control over the avalanche of new data items appearing by the
second, you will be contributing to the complexity that is the bane of
security.
What does scale is Accountability. In a free country,you don't have
to ask permission for much of anything, but that freedom is buttressed
by the certain knowledge that if you sufficiently screw things then up
you will have to pay. The economics of the access-control model of
information security do not scale; rather economics favor an
accountability model focused on the monitoring of information use
rather than the gatekeeping of information access. This means
surveillance of data use in the sense of being able to reconstruct how
information is used when it is used badly. This does not mean to throw
away our existing investment in access control, but further investment
in that will only produce inefficiency and a false sense of security.
We are, sadly if necessarily, making surveillance a commonplace of
physical security; it is no longer possible to live in a world without
cameras. We will have to, sadly if necessarily, make surveillance a
commonplace of cybersecurity. As you consider how to make these
dreadful choices, I suggest that the unit of observation be a datum,
not a person, that if a surveillance system has to protect the digital
world, that that surveillance be directed at data, not persons. If
anything, this is risk management applied to risk management.
Summary
We need a system of security metrics, and it is a
research grade problem.
The demand for security expertise outstrips the
supply,and it is a training problem and a recruitment problem.
What you cannot see is more important than what you
can, and so the Congress must never mistake the absence of
evidence for the evidence of absence, especially when it comes
to information security.
Information sharing that matters does not and will not
happen without research into technical guarantees of non-
traceability.
Accountability is the idea whose time has come, but it
has a terrible beauty.
Mr. Langevin. Mr. Saydjari.
STATEMENT OF O. SAMI SAYDJARI, PRESIDENT, PROFESSIONALS FOR
CYBER DEFENSE CHIEF EXECUTIVE OFFICER, CYBER DEFENSE AGENCY,
LLC
Mr. Saydjari. Chairman Langevin, Ranking Member McCaul,
members of the subcommittee, it is a pleasure to have this
opportunity to testify today on this matter of utmost national
importance.
I come to you as the leader of the Professionals For Cyber
Defense, a nonprofit group of recognized national cybersecurity
leaders advocating for sound U.S. cyber defense policy.
I have a written statement which, with your permission, I
would like to enter into the record. I will briefly summarize
it and look forward to responding to the committee's questions.
In 2002, more than 50 leading cyber defense experts signed
a letter, feeling compelled to warn President Bush of strategic
threat to our Nation from attacks to our information
infrastructure. Our message was simple. I am going to repeat
that message to you today. The U.S. faces a national strategic
threat requiring a national strategic response, and you can
help today.
First, to a strategic threat. The lack of a strategic
response must come, in our opinion, from a lack of belief in an
established strategic threat. Even an uncertainty and a
possibility of the strategic threat that we see demands
immediate action to resolve that uncertainty to move forward on
sound policy. Because of this, the Professionals for Cyber
Defense developed and vetted a simulated strategic attack
campaign against the United States to help establish the nature
and effect of such an attack.
Our findings are sobering. The U.S. is vulnerable to
strategically crippling cyber attacks from nation-state
adversaries. The level of devastation to our economy and to our
way of life is potentially disastrous. The ripping of our
social fabric will be on an order that we only glimpsed in the
aftermath of Hurricane Katrina. We will move from being a
superpower to a third world country practically overnight. We
are a Nation unprepared to defend ourselves against this
strategic threat and recover from it when it happens.
Therefore, the PCD recommends that the United States engage
in a national threat assessment immediately to verify our
findings and move forward. The critical IT infrastructure is as
legitimate a part of our territory as physical land. We depend
upon it now for our survival, just like land in the industrial
and agrarian ages. Cyberspace controls real-world critical
assets like power generators, power distribution, oil and gas
pipelines. The information age requires us to defend this
digital territory. Therefore, the government must provide for
the common defense of this new territory.
This is not a matter of big government versus small
government. It is not a matter of interfering or controlling
the private sector. The private sector openly has declared that
they desperately need the government's help against defending
against nation-state adversaries. There are a lack of
incentives for the private sector to solve this problem on
their own, just as there is a lack of incentive to solve this
problem to defend our land.
Second, the strategy response. An effective strategy
response is a multi-billion dollar national priority investment
run by the country's best expert focused on defensive
capabilities as soon as possible. This will require an
unprecedented level of collaboration between government and the
private sector. Think in terms of a national cyber militia,
where our private sector and government are working hand in
hand to defend our critical systems against nation-state
adversaries.
We must start now. The capabilities will take a minimum of
3-years to establish and will take beyond that to put into
effect. We cannot wait until we are in the middle of a disaster
to begin this development of these capabilities.
A program of this order requires a very, very large ante.
We estimate a $500 million ante to begin this program is
essential.
The organization is inherently multi-agency. Ultimately, we
will need a centralized national level, top talent, agile,
small special projects office to coordinate and run this effort
throughout this program.
Third, Congress can help today by doing three things:
First, support required funding levels. We are talking
about $50 million for the Department of Homeland Security R&D.
That is an order of magnitude off for the ante. We are in deep
trouble.
Second, advocate this initiative to agency heads in a
formal letter to motivate immediate discretionary investment to
begin to jump start this program right away.
Third, lead the way by commissioning blue ribbon panels and
special investigative committees to help establish momentum.
Inaction isn't an option for any of us who know the stakes and
are entrusted by the people to provide for the common defense
and to protect the future of this great Nation.
The PCD stands ready to help.
Thank you.
Mr. Langevin. Thank you, Mr. Saydjari.
[The statement of Mr. Saydjari follows:]
Prepared Statement of O. Sami Saydjari
Chairman Langevin, Ranking Member McCaul, and Members of the
Subcommittee, it is a pleasure to have this opportunity to testify
before you on an issue that is of utmost national urgency. I come to
you as the leader of the Professionals for Cyber Defense, a non-profit
group of recognized national cyber security leaders dedicated to
advocating for the development of a sound cyber defense policy for the
United States.
Summary. (1) The US is vulnerable to a strategically crippling
cyber attack from nation-state-class adversaries. Cyber space primarily
controls our real-world critical assets and is as legitimate a part of
our territory as physical land, thus the government must provide for
the common defense of this new territory. (2) A strategic multi-
billion-dollar investment run by the country's best experts can
mitigate this risk if we start now with $500 million. (3) Congress can
help today by supporting this funding level, advocating this initiative
to Agency heads in a formal letter to motivate immediate discretionary
investment, and leading the way by commissioning blue-ribbon panels and
special investigative committees to help establish momentum.
Imagine the lights in this room suddenly go out, and we lose all
power. We try to use our cell phones, but the lines of communication
are dead. We try to access the Internet with our battery-powered
laptops, but the Internet, too, is down. After a while, we venture out
into the streets to investigate if this power outage is affecting more
than just our building, and the power is indeed out as far as the eye
can see. A passer-by tells us the banks are closed and the ATMs aren't
working. The streets are jammed because the traffic lights are out, and
people are trying to leave their workplaces en masse. Day turns to
night, but the power hasn't returned. Radio and TV stations aren't
broadcasting. The telephone and Internet still aren't working, so
there's no way to check in with loved ones. After a long, restless
night, morning comes, but we still don't have power or communication.
People are beginning to panic, and local law enforcement can't restore
order. As another day turns to night, looting starts, and the traffic
jams get worse. Word begins to spread that the US has been attacked--
not by a conventional weapon, but by a cyber weapon. As a result, our
national power grid, telecommunications, and financial systems have
been disrupted--worse yet, they won't be back in a few hours or days,
but in months. The airports and train stations have closed. Food
production has ceased. The water supply is rapidly deteriorating. Banks
are closed so people's life savings are out of reach and worthless. The
only things of value now are gasoline, food and water, and firewood
traded on the black market. We've gone from being a superpower to a
third-world nation practically overnight.
We saw what happened to the social fabric when Hurricane Katrina
wiped out the infrastructure in a relatively small portion of our
country: chaos ensued and the impact lasted a long time. What would be
left after months of recovery from such devastation nationwide? Such
strategic cyber attack scenarios are plausible and thus worthy of
urgent attention. We are a nation unprepared to properly defend
ourselves and recover from a strategic cyber attack.
My purpose today is to make a case for congressional action to
support a major government initiative that could mitigate the risk of a
devastating strategic cyber attack against the US. To understand the
plausibility of such attacks without undertaking any action would be
unconscionable. Even uncertainty by government leaders regarding such
plausibility demands immediate action to remove the uncertainty and
enable responsible policy decisions. The only rational approach to
address a problem of this magnitude and scale is a concerted high-
priority government program on the order of the Manhattan Project.
Failure to embark on such a program now will have disastrous
consequences to our national interests sooner rather than later.
I will now review the case for action our group made in a letter to
President George W. Bush in 2002, highlight the true nature of the
national strategic threat in a realistic cyber attack campaign called
Dark Angel, outline the only reasonable strategic countermeasure in the
form of an urgent, high-priority, multi-billion-dollar national program
that we've dubbed the ``Cyber Manhattan Project,'' point to some recent
promising but woefully underfunded cross-agency analysis and planning
that affirms both the grave situation and the need for a national
program, and then I'll close with some recommendations on moving
forward.
Background. In 1939, Albert Einstein felt duty-bound to warn
President Franklin Roosevelt of a strategic threat to the country from
nuclear weapons and the need for immediate action. In 2002, more than
50 leading cyber defense experts similarly felt compelled to warn
President Bush of a strategic threat of a different kind, one to our
critical information infrastructure. On 11 September 2001, terrorists
used our air transport infrastructure against us and made a serious
impact on both our economy and sense of security. Against a strong
country such as the US, frontal attacks make little sense, but our
vulnerability to infrastructure attacks makes such attacks increasingly
likely.
The signers included a former Director of Central Intelligence, a
former Director of the National Security Agency, a former Director of
the Defense Advanced Research Projects Agency, and many of the nation's
leading scientists and engineers. We warned President Bush that (a) the
situation was grave, with nation-states such as China developing
serious offensive capabilities, (b) a national initiative with
priority, top talent, funding, and focus on par with the Manhattan
Project was urgently needed to create cyber defense capabilities in
close partnership with industry, (c) threading together components of
national exercises, results from accidental information system
failures, and actual cyber attacks, one could create devastating
scenarios of strategic damage to the US, and (d) that the private-
sector economy wouldn't solve the problem without government leadership
because of a lack of incentive to do so. Since we signed the letter,
little has changed with respect to the situation or the trend. It's
time to move forward.
A subset of the signers formed a group called the Professionals for
Cyber Defense (PCD) to engage in continuous advocacy. In summer 2002,
the PCD panel reviewed the President's draft National Strategy to
Secure Cyberspace. They found that the plan offered valuable advice to
counter lower-grade threats but that it had a fundamental flaw in its
unstated premise that there was no strategic national threat. In
response, we recommended that the government urgently initiate a
scientific process to establish the scale, gravity, and validity of the
national strategic threat of cyber war against our nation. We expected
that such a process would validate the repeated warnings from the
technical community in reports from the Defense Science Board, National
Academy of Sciences, and the President's Commission.
But in our dialogue with the government, we learned of two barriers
to aggressive action: (1) the perception that government investment
would require ``big government'' private-sector interference, and (2)
the case for national strategic vulnerability wasn't yet credible to
senior leadership. In retrospect, on the first issue, we failed to
realize that government leadership simply did not see cyber space as a
territory on which we deeply depend and that must be protected and
defended--rather, some people in leadership positions viewed it as an
optional digital playground of bits and bytes for exchanging personal
messages or looking at hobby information. But this isn't a matter of
``big government'' versus ``small government''; it's a matter of our
government stepping up to its constitutionally required duty to defend
the US against threats beyond the capabilities and means of the private
sector. We deeply understood the second issue, which is why we
advocated for an urgent national-scale analysis of the vulnerability as
the starting point for a program plan. In September 2002, the panel
decided to sketch a case for action in the form of a realistic
strategic cyber attack campaign against the US called ``Dark Angel.''
This sketch was intended to be a starting point because it could
demonstrate the problem's gravity.
The Threat: Dark Angel. What is the problem, and what is the
solution? For the problem, we must ask if a strategic national
vulnerability exists, what its scope is, and how bad ``bad'' can get.
Without understanding the detailed nature of the problem, the efficacy
of any proposed strategy is unknown. We must also ask why any proposed
national strategy will solve the problem, and what happens if it
doesn't. These seem like childishly simple questions, but the answers
have been elusive. Indications are that national economic devastation
is quite possible, and when we're in the middle of the disaster isn't
the time to start thinking about how to respond. Preparing for cyber
war will take in excess of three years and require infrastructure
instrumentation for critical computer systems, experienced cadres of
defenders who are well trained and exercised, control systems to
execute strategic responses, effective architectures to mitigate risk,
and a national program to create defensive capabilities. Thus,
understanding the problem is an immediate need.
Planning. The small PCD planning team included a campaign planner,
two experts in the financial sector, three in electrical power, and one
in transportation. We assumed only unclassified critical infrastructure
vulnerabilities. Our intent was to illustrate the damage a robust
campaign that used multiple attack paths could cause and to create a
plan with sufficient detail to convince experts in the domain. The plan
took roughly 30 days to create. We assumed the adversary had three
years of preparation, $500 million, and 30 days to actually execute the
attack. The attack campaign's goal was to destabilize the US and
depress the economy with attacks on critical infrastructure, thus
reducing our ability to project military power, depleting our will to
fight, and creating panic and distrust in the government.
Our strategic campaign objectives included crippling rail
transportation, rupturing oil and gas pipelines with improper control
(for example, with cyber attacks similar to the one on the Soviet
Trans-Siberian pipeline causing a three kiloton explosion, as described
in ``At the Abyss'' by Thomas Reed), and creating widespread power
outages by destroying hard-to-replace generators and power-line
transformers with improper computer control commands. We also simulated
attacks on financial services sectors, thus creating mass confusion in
transaction settlement systems, flooded 911 systems with computer-
controlled false alarms to create widespread panic, and disabled
Internet service by performing denial-of-service attacks on the 13 main
Domain Name Servers (as has already been partially done in actual cyber
attacks).
In the simulated campaign, we spoofed attack attribution when
possible to focus attention in the wrong direction; used lethal first
strikes (for example, by hitting first responders and backups before
hitting primary cyber targets); used a rolling attack barrage to
interfere with recovery processes; delayed attacking instruments, such
as the Internet, until that means was no longer needed in the campaign;
bought cyber mercenaries and insiders as needed to gain capabilities
and access; used non-cyber (physical) attacks on ``tough'' targets as
needed; used psychological operations to create distrust in
infrastructure and manipulate public opinion; and hampered the military
by disrupting civilian re-supply chains.
Our simulated attacks were vetted with experts in each of the key
critical infrastructure domains. The essence of the plan and its likely
effects were verified. There was some uncertainty about the
consequences of some attacks--even now--but this was due to a lack of
knowledge among the entire community to fully assess such consequences.
It would be hubris to think our adversaries don't already have a plan
in place that's substantially better than our brief sketch or that
their capabilities to execute such an attack aren't improving.
Follow-on. A proper national strategic threat assessment would
parallel that of Dark Angel, and would involve top industry experts and
business leaders, mix in military campaign planners, and mix in
economists, policy makers, and others as needed. Sharing across
industry should be encouraged and rewarded. From a management
perspective, the assessment should carry presidential authority and
priority. There should be three separate teams: one for planning and
completing a concrete plan, one to execute the plan to the extent
needed for demonstration purposes, and one to review the results for
validity.
The assessment must start from the premise built into Dark Angel:
that cyber warfare will be economic and social warfare. Diagnosis of
the source of vulnerabilities must be included and reflect that the
organization and design of our production systems will often be more
important than cyber defense technology in determining the nature and
extent of the destruction. What to defend and what kinds of damages to
prevent are not self-evident without such an assessment.
For illustrative purposes, we estimate the resources needed for six
critical infrastructure domains would take about $70 million, 300 top-
talent experts, and 9 calendar months. The final report would be a
definitive estimate of our true national strategic vulnerability to
cyber attacks, a compelling case for action, and the basis of a
prioritized program plan.
Countermeasure: Cyber Manhattan Project. As part of our dialogue
with the government in 2002, we elaborated on the proper solution to
the strategic vulnerability sketched out by our Dark Angel analysis.
Cyber war defense requires orders of magnitude more government
involvement and resources to avoid overwhelming national damages from
strategic attacks. We recommended that the government (1) step up to a
strong defense role against serious attacks, (2) focus on countering
strategic attacks that have real-world effects, (3) develop a top-down
architecture and engineered approach to the defined problem, (4)
acknowledge that current technology is insufficient to defend against
cyber war, and (5) divide the cost burden between the owner (to protect
critical private cyber assets) and the government (to protect the
integrity of the national commons).
As mentioned earlier, we chose the name ``Cyber Manhattan Project''
to reflect the urgency, priority, focus, top-talent, and funding levels
needed. We acknowledge that aspects of the analogy are inapt, such as
the fact that (1) there is no single, easily measurable artifact (such
as a bomb), (2) a broad spectrum of talent and organizations must be
involved, (3) much of the work must be conducted without classification
constraint, and (4) once an initial capability is achieved, a continued
investment will be needed to maintain our cyber defense's
effectiveness. We sketch the program below.
Vision. We must rapidly overcome our nation's vulnerability to
coordinated strategic cyber attacks from serious enemies.
Project Description. We need an aggressive, goal-directed, high-
priority, national program to address the high-level threats that
endanger the national well-being. To do this, we must engage the
brightest scientists, business experts, and engineers, and provide them
with adequate resources. To guide the program with strategic
objectives, we need a top-down architecture that establishes concrete
cyber defense capabilities on a specific timeline, including near-term
capabilities within three years.
Capabilities. Some cyber defense capabilities to include are as
follows: (1) capability to create system resiliency and quickly recover
from inevitable partially successful attacks; (2) a national cyber
Command, Control, Communication, and Computer Intelligence,
Surveillance, and Reconnaissance (C4ISR) system to measure and control
mechanisms at multiple echelon levels; (3) a national threat assessment
capability to drive decisions at some ``required'' level; (4) cyber
firebreak mechanisms and architectures to slow down attacks and reduce
potential damage; (5) capability to gather intelligence and inject
uncertainty through strategic deception; (6) capability to model and
simulate the enemy, thereby honing our defenses before incurring
damaging strategic cyber attacks; and (7) capability to identify and
understand available and acceptable responses from technical,
strategic, legal, economic, and political perspectives.
Urgency. Major potential adversaries are actively pursuing cyber
war capabilities, which indicates the increasing probability of future
cyber campaigns. Moreover, (a) current cyber defenses and best
practices are ineffective, (b) active measures to shut down our
adversaries' abilities to attack through physical access will drive
them to cyber space, and (c) we face potentially greater vulnerability
and lethality from combined cyber and physical attacks. Finally,
developing a defense to this threat is a multiyear effort, so we can't
wait until we find ourselves suffering in the midst of our first major
strategic attack campaign.
Priority. A major initiative on the order of the Cyber Manhattan
Project is the right path to address our current situation. The
offensive threat is growing, so defense must be fielded at a faster
rate. A top-down approach with a driving architect can address the
problem and achieve the requisite objectives, but bottom-up efforts,
even if coordinated, leave gaps because there's no ownership of key
parts of the problem. Cyber defense mechanisms must integrate into a
coordinated system, and cyber defense operations must comprise a fully
integrated defensive force. For success, the creation of national cyber
defense capabilities must be a national funding priority. Can you
imagine the original Manhattan Project succeeding without such a focus?
Feasibility. Not only is the creation of national cyber defense
capabilities critically urgent and important, it's also feasible. (1)
Technically, many effective defensive technologies exist but are in
research stages and must be transitioned to operational use; some
already have limited field testing, and others already exist to address
broad classes of novel attacks. Moreover, the required computational
resources for intensive activities such as correlation of attack and
modeling/simulating attack strategies and tactics are available today.
Ongoing research sponsored by the likes of NSA, NSF, DOD, DNI, DHS, and
others is beginning to address additional hard science problems. (2)
Economically, we can make a national business case for investing in a
program intended to avoid the expected financial losses from strategic
cyber attacks and ensure the proper public-private sharing of the
burden. (3) Operationally, we can manage the complex infrastructure
though judicious use of automation with a capable cadre of defenders.
Through a combination of reasonable fire-code-like cyber security
standards, improved operational guidance, and trained/experienced
personnel, we would also be able to contain mission and cost impacts in
the short term while we develop new capabilities. (4) Politically,
public awareness of the threat is likely to make needed investments and
standards acceptable. Industry is increasingly aware that nation-state-
level attacks are a concern beyond their current ability to handle, yet
they threaten business continuity. With proper financial incentives and
partnering for workable solutions, industry is likely to openly embrace
government involvement and protection. (5) Finally, from a schedule
perspective, a phased rollout of capabilities based on threat
prioritization and available technologies is also feasible. Success is
certainly not assured, but the alternative is to begin radically
reducing our dependency on computing systems, which would seriously
degrade our national competitiveness and suppress economic growth. The
cyber vulnerabilities in our infrastructures have become deeply
embedded and widespread through the economic forces that drive
individual companies to reduce costs by adopting the most widely
available and interoperable technologies. It won't be easy to develop a
cyber infrastructure that can resist strategic attacks--it will require
short-term actions as well as a long-term plan and a willingness to
keep that plan in focus over a number of years.
Plan of Action. We recommend assigning a government lead
responsible for creating a plan. The PCD offers to work with this lead
and recommends a three-month deadline for developing a ``blueprint'' to
launch the project, including technical and program management aspects.
We also recommend jumpstarting a multiyear program now with as much
seed funding as possible.
The PCD hasn't worked out a full recommendation for how a Cyber
Manhattan Project, which would inherently involve multiple agencies,
ought to be organized and managed. A few points of consensus, though,
appear to be emerging. (1) Distributing a surge of funding to the
myriad bureaucracies that currently fund cyber defense won't work in
the long run. Each bureaucracy pulls in a different direction, making
focused investment nearly impossible, although a jumpstart in 2007/2008
might have to start this way out of sheer practicality. (2)
Centralizing funding and government-wide responsibility in one existing
department or agency with its own mission will likely cause the funding
to be spent by that bureaucracy's priorities, to the detriment of
national interest. (3) Creating a whole new department or agency might
fall into the too-hard-to-do pile, given the tremendous distractions
and delays involved (as we've seen with the startup of the Department
of Homeland Security).
Eventually, what we need is a centralized, light-weight, high-level
controlling body to create a focused effort on national cyber defense
capabilities. One thought has been to create a special projects office
accountable to and operating with the authority of the White House,
with an elite staff of 200 people, at least half of the overall program
budget, and some purview over the spending of the other half
distributed and executed by existing organizations.
Recent Developments. Recent activities tend to echo and affirm the
PCD's earlier findings. In November 2006, in response to concerns of
inherent computer system vulnerabilities and escalating threats, more
than 60 experts in system security, processor design, operating
systems, programming languages, networking, and applications from
diverse backgrounds in academia, government, and industry met to
consider past, current, and possible future approaches to building
systems with improved security. Findings from this Safe Computing
Workshop included the following: (1) attackers rule, disasters are
likely; (2) short-term measures are essential but insufficient; (2)
market forces won't change the balance; (3) usability and manageability
must be part of the solution; (4) new technology can catalyze major
changes; and (5) only a national initiative will make a real
difference.
The workshop participants also concluded that the timing of such an
investment is particularly good now because (1) significant advances in
technology have dramatically increased hardware processing, memory, and
communication capacity; (2) there's a growing understanding of the
problem among the public and government leadership as everyday cyber
attacks like spam, phishing, and identity theft become increasingly
painful; (3) industry's interest in cyber security continues to grow as
the community becomes more adept at making a business case for
improvements; (4) escalating attacks and damages are increasing across
the globe; (5) major software vendors are willing to delay the release
of their products for more than a year to forestall security
embarrassments; and (6) without a major change in direction,
adversaries will be able to exploit current weaknesses in US cyber
security and could deal a critical blow to our country's major
industrial sectors, such as banking, energy, and telecommunications.
The workshop participants found a compelling and urgent need to
dramatically reduce the vulnerability of the national information
infrastructure to attack, and that major, strategic investments could
significantly reduce our vulnerability over a five-year period.
Closing Remarks.
Smoking Gun. Some of you might think, what's the rush? Where's the
smoking gun--the indication of a major assault on US cyber
infrastructure? Surely, it's coming, and it's no doubt already in its
planning stages. We suggest three reasons for why this is so. First,
strategic long-term damage requires substantial planning and very well-
timed execution. Creating the capabilities and placing the required
assets (such as insiders) takes time, certainly years. Second, when
such a cyber attack weapon is created, it's in some sense a one-time-
use strategic option. One wouldn't use it lightly, nor would one want
to tip one's hand about it until it's really needed: such weapons may
well be deployed already, and we wouldn't know it (perhaps a sleeper
cell of insiders and/or malicious software embedded in our critical
infrastructure). Finally, our current cyber infrastructure offers a
wealth of highly valuable knowledge (such as advanced research
results). As adversaries conduct espionage, they're also mapping our
cyber space and gaining great experimental and training experience that
will enable future strategic attacks. It's in the interests of our
adversaries to preserve their upper hand for as long as possible and
keep tapping into these important attributes. Moreover, such nation-
state network exploitations are becoming increasingly obvious to the
point that the mainstream press regularly covers them.
Secrecy. We don't advocate that a Cyber Manhattan Project be
shrouded in secrecy: doing so would be unnecessary and deleterious to
the program goals. The nation's best minds must work on this difficult
problem, and many of them are to be found outside government in
academia and industry. Excluding those minds by making the program
secret would only decrease our chances of success. Obviously, it makes
some sense to maintain the element of surprise about the details of
some of our planned defenses, but these should be carefully thought out
and very limited in scope. A design that counts on its own secrecy to
succeed isn't a robust design at all: we all know how fleeting secrets
can be.
Stakes. But what if we don't do this? Ladies and gentleman, based
on the vetted Dark Angel scenarios, we could compromise our country as
we know it if we make a misstep today. Inaction isn't an option for any
of us who now know these stakes and are entrusted by the people to
provide for the common defense and protect the future of our great
country. Thank you.
Mr. Langevin. Dr. Maughan.
STATEMENT OF DOUGLAS MAUGHAN, PROGRAM MANAGER, CYBER SECURITY
R&D, DEPARTMENT OF HOMELAND SECURITY, SCIENCE AND TECHNOLOGY
DIRECTORATE
Mr. Maughan. Chairman Langevin, Ranking Member McCaul,
members of the subcommittee, thank you and good afternoon.
Today, I will be sharing with you information on the
cybersecurity research and development program in the
Department of Homeland Security Science and Technology
Directorate. I also will outline for you critical areas where
new research and development efforts are needed. Details of the
Directorate's program are included in my written testimony. I
will provide a brief summary.
The program's mission is to drive cybersecurity
improvements in existing and emerging technologies; discover
solutions to detect, prevent and respond to attacks on our
critical infrastructure; and deliver new, tested solutions for
cybersecurity threats, making them widely available to all
sectors.
Unlike other government programs, we cover all phases of
the R&D lifecycle, not just research, but research,
development, testing, evaluation and transition. Because our
research is unclassified, we produce solutions that can be
implemented for our customers in both the public and private
sectors. We aim for results that can have impact in every home
and business in the U.S. and throughout the world because cyber
threats affect everyone.
Consider the following: Cybersecurity breaches have real
economic consequences. Internet users who shop online spend an
estimated $8 billion per month. But according to a recent
Consumer Reports survey, 86 percent of American internet users
have changed their behavior due to fears of online theft; 25
percent have stopped shopping online altogether for that
reason.
A 2005 Cybersecurity Industry Alliance study found that 65
percent of American voters indicated that the government needs
to do more to protect our information and systems from
cybersecurity threats. Worldwide cyber attacks were estimated
by the Congressional Research Service at a cost of $226 billion
in 2003. The cost impact of these attacks is most certainly
higher today.
The DHS Cybersecurity Research and Development Program
budget totaled $13 million in fiscal year 2007. The President
has requested $14.8 million for fiscal year 2008. I would like
to share with you some positive results that we have
accomplished.
We have funded small businesses and universities to solve
near-term cybersecurity problems, such as malicious code
detection, insecure wireless networks, open source software
vulnerabilities and identity theft.
We have funded research that has led to more than 10 open
source and commercial products in the past 3 years alone.
Examples include secure thumb drives, root kit detectors and
security solutions for process control systems. We have brought
together entrepreneurs, venture capitalists and system
integrators to speed the transition of these innovative
cybersecurity solutions for commercial and government use.
We have created a cybersecurity testing environment
comprised of a test network and test data sets containing real
traffic data to support the research community.
And we have led an international effort to advance the
deployment of critical solutions required to secure the
Internet infrastructure as called for in the National Strategy
to Secure Cyber Space.
We need to continue our efforts to bring these important
cybersecurity solutions forward, but more is needed. The DHS
Science and Technology Cybersecurity Program, in concert with
our customers, has identified five research areas as priorities
which we will continue to address as we face the future.
We need to develop more secure versions of basic Internet
protocols and architectures to ensure that the Internet works
safely the way users expect it to.
We need to create new ways to detect and contain attacks
and develop resilient systems and detect and mitigate insider
threats.
We need to build research infrastructure and tools to
support cybersecurity research and development efforts.
We need to find new technologies to reduce the
vulnerabilities in our process control systems that underlie
our Nation's critical infrastructure.
And we need to develop trusted systems and the metrics to
assess them.
Mr. Chairman and members of the subcommittee, the good news
is we are making progress. The Directorate's research and
development results show promise, and I look forward to working
with you to address the security needs of the Nation's critical
infrastructure.
Thank you. I look forward to answering any questions you
may have.
[The statement of Mr. Maughan follows:]
Prepared Statement of Dr. Douglas Maughan
Chairman Langevin, Ranking Member McCaul and Members of the
Subcommittee, thank you and good afternoon. Today, I will be sharing
with you three important aspects of our work in cyber security research
and development in the Department of Homeland Security (DHS) Science
and Technology (S&T) Directorate, including our efforts to:
Drive security improvements in existing technologies
and emerging systems.
Discover solutions to detect, prevent and respond to
cyber attacks on the Nation's critical infrastructure.
Deliver new, tested solutions for cyber security
threats and make them widely available to all sectors through
technology transfer and other methods.
The S&T Cyber Security R&D goes through the full R&D lifecycle--
research, development, testing, evaluation and transition--to produce
unclassified solutions that can be implemented for our customers in
both the public and private sectors. Therefore, we are able to move
these solutions from the lab to real life, so they reach the U.S.
businesses and citizens who need them to secure their networks. It
means that the results of our research can have an enormous impact in
every home and business in the United States, as well as throughout our
government and the world. In the past three years alone, the DHS
Science and Technology Directorate has funded research that today is
realized in more than 10 open-source and commercial products that
provide capabilities such as: secure thumb drives, root kit detection,
worm and distributed denial of service detection, defenses against
phishing, network vulnerability assessment, software analysis, and
security for process control systems.
Cyber threats pose an ever-growing risk to our national and
economic security. We face enormous challenges in our ability to meet
or even anticipate those threats. Today, I hope to describe briefly for
you: the scope of the problem; and the positive steps we are taking to
drive, discover and deliver new solutions.
The events of September 11, 2001, made clear that the security of
our Nation and our economy are intertwined. The majority of government
communications utilize private-sector networks, including critical
infrastructures--such as information technology, communications,
financial services, electricity, and oil and gas systems. These
networks have proven interdependencies that are critical to response
capabilities as well as business operations. The systems of these
sectors have converged and are interconnected. For example, if the
electrical grids fail, that failure impacts the communications systems,
which in turn can hamper financial networks.
The Internet connects all other networks, including our Nation's
critical infrastructure. It has become the central nervous system for
our government, our citizens and our industries. When it is attacked,
the effects can ripple far and wide. Although the Internet was
developed to provide ``essential minimum communications'' in the event
of a nuclear attack, it was not designed with security in mind. Thus,
the technology that is deployed over most of the Internet today has
vulnerabilities that can be exploited, endangering all the connecting
networks, including our critical infrastructures.
Beyond the Internet, few of the technologies we use every day are
adequately protected against malicious attacks. Cell phones, PDAs, and
wireless networks are vulnerable, as are the supervisory control and
data acquisition (SCADA) systems underlying our critical
infrastructure. Attacks on these technologies have forced us into a
defensive posture, and the financial costs are significant. Attackers
can reach our business and government systems through the maze of
networks connected by the Internet.
A 2004 Congressional Research Service (CRS) report stated that
cyber attacks on publicly traded firms resulted in losses of 1 percent
to 5 percent on the firms' stock price in the days following an attack.
For the average New York Stock Exchange company, this means shareholder
losses in the range of $50 million to $200 million. CRS reported that
total losses worldwide in 2003 attributed to viruses, worms, and all
other hostile digital attacks were $226 billion. These attacks can come
from rogue actors (such as script kiddies, disgruntled employees, and
organized crime), terrorists, insiders, and other nation states.
But it is not just companies and governments at risk: Our citizens
also are vulnerable. Government action can help protect U.S. consumers
who, in many cases, cannot adequately protect themselves from threats
that come from our cyber infrastructure. Countering these threats
requires the deployment of new technologies across the global
infrastructure.
Americans make extensive use of the Internet. March 2007 global
statistics indicate there are more than 210 million Americans--70
percent of our total population--using the Internet. On their private
computers, our citizens are targeted by viruses, worms, and phishing
schemes. Their computers may be used as launching pads for attacks
against other systems, unbeknownst to the computer owner. To date, more
than 150 million records containing personally identifiable information
have been exposed since January 2005, according to the Privacy Rights
Clearinghouse.
According to a 2005 Consumer Reports survey in the U.S., 86 percent
of Americans who go online have made at least one behavior change due
to fears about online theft. 29 percent have cut back on shopping
online, and another 25 percent have stopped shopping online altogether.
A 2006 survey from the Cyber Security Industry Alliance (CSIA) found
that Internet users who do shop online indicate that they spend an
average of $116 per month per person--an estimated $8 billion per month
in total--but that half of all users avoid making purchases because of
fear of identify theft or compromise of financial information.
Indeed, citizens want the Federal government to bring forward cyber
security protections. A 2005 survey of U.S. voters--both Internet users
and non-users--conducted by CSIA found that respondents look to the
U.S. government to help with cyber security issues. Sixty-five percent
of the respondents indicated that the government needs to do more to
protect information and systems.
In fact, the Department of Homeland Security's Science and
Technology Cyber Security program serves all of these customers, which
include both DHS internal components and private sector entities: Cyber
Security and Communications (which includes the National Cyber Security
Division and the National Communications System), U. S. Secret Service,
DHS Chief Information Officer (CIO), Internet infrastructure owners and
operators, critical infrastructure providers, and the information
security research community. The Directorate leads the government's
charge in funding cyber security research and development that results
in deployable security solutions, as directed by the President in the
National Strategy to Secure Cyberspace. Our research and development
funding is targeting the critical problems that threaten the integrity,
availability, and reliability of our networks. We provide solutions and
research resources that advance our understanding of cyber security
risks. Our goals are:
To protect our national and economic security
interests and secure our homeland.
To enable the government, industry, and citizens to
make better-informed decisions about cyber security risks.
To provide the resources needed to counter and
mitigate these risks.
The United States played a formative role in the Internet's
creation, and is home to ten of the thirteen root servers that control
the communications flowing over the Internet. However, today's security
vulnerabilities cannot be addressed in isolation. Today, there are 243
countries connected to the Internet and approximately 1.2 billion
online users worldwide. It is a global problem that affects
governments, businesses, and citizens. To get this important work done,
the S&T Cyber Security R&D program carefully collaborates with private
industry, Federal agencies and other governmental entities, and
private-sector partners in other nations, reflecting the truly global
nature of the Internet.
There are legal issues and international coordination issues that
need to be addressed, but there are also complex technical problems
that need to be solved. The price tag for this research and development
is high, but it is minimal compared to the cost of cyber attacks today.
Let me restate for the members of the Subcommittee that worldwide cyber
attacks were estimated by CRS at a cost of $226 billion in 2003. The
cost impact is most certainly higher today. The Department of Homeland
Security's Science and Technology Directorate's cyber security research
and development budget totaled $13 million in FY 2007 and the President
has requested $14.8 million for Fiscal Year 2008.
Today, I'm going to discuss three important areas where we are:
Driving security improvements to address critical
weaknesses in the Internet's infrastructure
Discovering new solutions for emerging cyber security
threats, by incubating ideas and innovation in safe testing
environments and public-private partnerships
Delivering new technologies tested in a real-world
environment and making them widely available for real-world
users in all sectors
I also will describe for you those research areas identified in
concert with our customers that are ongoing priorities which we will
continue to address in FY2007, FY 2008 and beyond:
Driving Security Improvements to Address Critical Weaknesses
The Department of Homeland Security's Science and Technology
Directorate is leading efforts to secure two of the Nation's major
technology vulnerabilities: security weaknesses in the Internet's
domain name system, or DNS, and vulnerabilities in the Internet routing
system. Attacks against these two parts of the Internet infrastructure
are particularly insidious because computer users cannot detect them.
Attack traffic is estimated to have skyrocketed 150-fold since 2000.
Both domain name system and routing vulnerabilities can deny
service to small or large portions of the Internet, make tracking and
tracing Internet communications very difficult, or allow communications
to be redirected without the user's knowledge. In the dot-com and dot-
net domains alone, domain name queries are made an average of 24
billion times a day, yet Internet users have no guarantee that they
will reach the Web site they want when they enter its address in a
browser. Symantec's most recent Internet Security Threat Report notes
that, in the first six months of 2006, spam made up 54 percent of all
monitored e-mail traffic. Much of that spam takes advantage of
weaknesses in the routing system, and uses it to mask spammers'
identities, making it difficult, if not impossible, to track them down
and prosecute them.
U.S. government leadership in addressing these critical
vulnerabilities is essential, and the President's National Strategy
calls on DHS to drive the efforts to bring solutions forward. By
working in a collaborative effort across Federal agencies, private
industry, and global Internet owners and operators, the DHS Science and
Technology Directorate has made progress toward addressing these
problems. In cooperation with NIST and the Department of Commerce, our
Directorate leads the effort to develop domain name security extensions
(DNSSEC), and we work with international counterparts and key technical
groups to develop improvements to the standards that govern addressing
and routing.
Both of these infrastructure security problems have, or soon will
have, solutions driven by our government's leadership. The remaining
challenge lies in convincing the many owners and users of the Internet
to deploy them, from private industry and foreign governments to our
own state, local and federal agencies in the U.S. New requirements
under the Federal Information Security Management Act (FISMA) call for
DNS security extensions to be deployed across all federal agencies and
their contractors. A few other countries, notably Sweden, have already
deployed the important DNS security solution.
The private sector also is starting to follow the government's
lead. Two major corporations working in software and information
security also have announced plans to include DNS security extensions
in their products going forward. Microsoft, which supplies the
operating system for the vast majority of the U.S. government's desktop
computers, will include the new DNS security protocols in a forthcoming
upgrade of its software. VeriSign also has announced that it will
include the DNS security protocols as part of an expansion that will
enable it to handle more than four trillion domain name system queries
per day. Many more government agencies and industries must take similar
steps if we are to secure the Internet infrastructure.
The government has a special role to play in coordinating the
deployment of these solutions. The S&T Cyber Security R&D program is
positioned to carry this work forward. Building on our research and
development efforts, the government can play an even greater leadership
role by taking steps to ensure the government-wide deployment of DNS
security extensions and secure routing technologies, when available.
Discovering New Solutions for Emerging Cyber Security Threats
We cannot focus solely on known problems. One of the most important
aspects of cyber security R&D involves understanding new threats and
risks, and discovering solutions that will help us protect our Nation's
cyber infrastructure. Because the research we conduct is unclassified,
it can be deployed by the private sector. The S&T Cyber Security R&D
program funds two efforts that provide a safe environment for cyber
security research. Using small business innovation research funding and
other programs in our Directorate, we also provide funding that helps
bring forward the next generation of cyber solutions so they can be
adapted for wider use against emerging threats. With more than 30 small
business innovation research grants in progress today, as well as other
funds, we are incubating ideas that emanate from small companies and
devising solutions for emerging problems that will affect major
sectors.
The need to create, test, and learn from potential threats poses a
problem in itself. We want to test threats to the Internet, but if we
conduct such R&D testing on the actual Internet, we could inadvertently
put it at risk. To provide scientifically rigorous testing for next-
generation cyber defense technologies, the DHS Science and Technology
Directorate funds a cyber security testing environment, comprised of a
test network, and test data sets containing real-traffic data.
The network, called the Cyber Defense Technology Experiment
Research Testbed Program, or DETER, offers cyber security researchers a
way to run experiments on a secure ``virtual Internet,'' keeping the
Internet safe. This testbed was jointly funded with NSF and now more
than 50 organizations from more than 20 states--which includes major
research universities, national laboratories and high-tech companies--
are using the DETER test bed. The test bed began with 200 systems, and
has been increasing by 200 per year with a goal of 1,000 systems spread
across six sites by FY09.
In addition to a test network, researchers need data sets to use
for testing their solutions. These data sets, however, have not
existed, impeding effective testing of potential technologies. For
example, the most widely used data source today was created in 1998 by
the Defense Advanced Research Projects Agency (DARPA). Traffic data
that is nine years old cannot be used to analyze today's attacks,
viruses, malicious code, and traffic patterns.
The S&T Cyber Security R&D program created and funded the Protected
Repository for Defense of Infrastructure Against Cyber Threats, or
PREDICT program, to serve as a repository for a collection of datasets
that can be used for testing new ideas and solutions. PREDICT provides
datasets for information security testing and for the evaluation of
maturing network technologies, to help advance them toward commercial
development. The PREDICT data repository also is designed to hold
datasets which can be collected from private companies, without
violating their proprietary concerns, for sharing with network security
researchers. The PREDICT program has taken groundbreaking steps to
ensure that data privacy is protected, including reviewing the project
with major privacy organizations.
As I noted earlier, another critical area of focus for the DHS
Science and Technology Directorate is the development and deployment of
the next generation of cyber security technologies that we need if we
are to effectively face emerging threats to our Nation's critical
infrastructure. We solicit research proposals for new technologies,
prototype technologies and mature technologies, so that our investment
yields solutions that are poised for commercial adoption. Under the
first round of this research funding effort, we awarded $13.8 million.
The $13.8 million funded projects in 12 states: California, Delaware,
Georgia, Massachusetts, Maryland, Michigan, Minnesota, New Hampshire,
New Jersey, New York, Texas, and Virginia.
Let me give you some examples of projects we've funded in this
area:
In California, Stanford University researchers are
identifying and fixing serious bugs in open source code for
freely available software. Widely used, open source software
makes up a large part of the Nation's cyber-infrastructure, and
this effort has lead to tools that are available through a
commercial company named Coverity, located in San Francisco and
Boston.
In Ann Arbor, the University of Michigan's researchers
are working on a secure crisis response system using handheld
devices. Using low-cost disposable handheld devices, first
responders will be able to have a secure mobile coordination
and syndication channel--a lightweight means for interagency
communication and coordination using industry-standard wireless
and cell phone technologies, while keeping data transmission
secure. This project partners with Lucent Technologies for
commercial deployment.
At Dartmouth College, researchers are analyzing
wireless traffic to detect and respond to attacks on a WiFi
network. The project is working with Aruba Networks of
Sunnyvale, California, a very large wireless vendor in the
United States, to develop and deploy an operational prototype
and evaluate it with real-time users.
Additionally, we are partnering with the financial sector to assess
the economic impact that a cyber security attack might have on
individual enterprises, and developing tools to help financial
companies assess and manage the risks that such a disruption of service
could create.
Working with companies like Citigroup and Pershing LLC, a brokerage
subsidiary of the Bank of New York, we have created a prototype of a
risk management tool for the finance sector. It is designed to help
create a computer simulation of a financial enterprise and its value
chains, and how they interconnect with other institutions. Once it is
finalized, the tool will allow them to create and run disruption
scenarios tailored to their business operations, using their own
proprietary data as well as generic data for the rest of the financial
sector. In this way, they can find out specifically how a cyber
security event or attack will affect their business, using real-time
sector data while protecting their companies' proprietary data.
I want to underscore the special role that government funding has
played in developing this prototype. No single financial company would
build such a tool and share it with competitors; however, because of
support from our Directorate, the entire financial sector will be able
to assess and protect itself against emerging cyber security threats,
protecting our Nation's critical infrastructure.
Delivering New, Tested Technologies Widely Available for All Sectors
New cyber security solutions do not appear in products
automatically. Technology transfer from the lab to the marketplace is a
vital and unique aspect of our Directorate's cyber security R&D effort.
The S&T Cyber Security R&D program extends beyond knowledge and the
proof of whether security solutions are feasible. Based on this
foundation of rigorous research and development, we create public-
private partnerships, acting as a catalyst to deliver new, tested
technology solutions for cyber security threats and make them widely
available for use in all sectors.
One important test we have conducted focused on handheld wireless
devices, like the BlackBerry and other mobile data communications
devices. These devices are expected to proliferate within government
agencies. According to a 2005 survey in Government Computing News, 40
percent of all government managers report that they use some form of
handheld wireless device. Hundreds of thousands of these devices are
currently employed in government business, yet today, most mobile data
architectures cannot sufficiently assure high-level government
security.
To address those issues, and to identify the needs in
infrastructure protection and border security, we conducted an
experiment under the bilateral Public Security Technical Program
between the United States and Canada. It is just one of many efforts by
the DHS Science and Technology Directorate to evaluate technologies in
a real-world environment and pass on the results to real-world users.
Our research was looking for new technology for mobile data encryption
across the US-Canada border, to learn whether additional security
measures would slow down communications across the borders, and to help
first responders tackle their tasks efficiently while keeping their
messages secure. We tested four products of interest, including the
BlackBerry, and learned a great deal about what does and doesn't work,
particularly situations in which messages were delayed, or data were
not transmitted.
Another important public-private partnership is Project LOGIIC,
which stands for Linking Oil and Gas Industry to Improve Cyber
security. The goal is to reduce vulnerabilities in the oil and gas
process control system environments. The first demonstration under this
project showed how to correlate and analyze abnormal events to identify
and prevent cyber security threats.
Project LOGIIC is a model for government-industry technology
integration and demonstration efforts to address critical research and
development needs. The oil and gas industry contributed the
requirements, operational expertise, project management, and product
vendor channels. DHS provided the national security perspective on
threats, access to long-term security research, independent researchers
with technical expertise, and testing facilities. Technology pilot
deployments under this program were launched in June of 2006. A
planning meeting for the second phase of the LOGIIC partnership took
place in March of this year.
Our Directorate also convenes a group called the Identity Theft
Technology Council, which meets three times a year to bring together
government, venture capital firms, financial sector representatives,
academics working in identity theft, and entrepreneurs. Together, we
discuss problems, research issues, available technologies, and stay
abreast of emerging threats and new opportunities. As a result, venture
capital firms and the companies that they fund can connect with
government and larger private-sector entities to move emerging security
solutions forward. The Council also works closely with the Anti-
Phishing Working Group, and has issued two reports: one on phishing and
one on malware.
To help technology move out of government research and development,
we have sponsored three different types of transition forums:
At the System Integrator Forum, researchers funded by
the DHS Science and Technology Directorate were provided an
opportunity to demonstrate their technology to an audience of
major system integrators, including Perot Systems/EDS, Northrop
Grumman, and General Dynamics, all of whom responded
enthusiastically.
The Emerging Security Technology Forum provided an
opportunity for commercial developers to demonstrate their
technology to an audience of government early adopters. Our
Directorate evaluated 24 commercial technology products to
defend against distributed denial of service and worm attacks,
and selected 12 for presentation to an audience of government
and industry CIOs and potential customers.
Finally, the IT Security Entrepreneurs Forum--jointly
sponsored with the Kauffman Foundation--provided small
businesses and entrepreneurs an opportunity to learn value
propositions and business plan development from the venture
capital community and how to open doors into government
procurement channels. Chief information officers attended from
companies like Sun and Oracle.
The impact of these forums cannot be overstated. They are unique
within the federal system. We bring researchers directly to the private
sector, so they can demonstrate their technologies in front of more
than 100 companies at a time. As I mentioned earlier, this has led to
more than 10 commercial cyber security products--real cyber security
solutions that can be widely used by government, industry and citizens
around the world. These forums assist projects funded by our Science
and Technology Directorate to transfer technology to larger,
established security technology companies. Finally, they also help
commercial companies provide technology to DHS and other government
agencies.
Driving, Discovering and Delivering Cyber Security Solutions: The
Path Forward
In the last seven years, more than 20 reports from such entities as
the INFOSEC Research Council, the National Science Foundation, the
National Institute of Justice, the National Security Telecommunications
Advisory Committee, the National Infrastructure Advisory Council, the
National Research Council and the President's Commission on Critical
Infrastructure Protection have urged the government to do more to
drive, discover and deliver new solutions to address cyber
vulnerabilities. More recently, academic organizations, such as the
Computing Research Association, and industry groups, such as the Cyber
Security Industry Alliance and the Internet Security Alliance, also
have called for increased funding for cyber security research and
development. In addition, the Federal Government has recently produced
the Federal Plan for Cyber Security and Information Assurance Research
and Development, which includes cyber security R&D priorities of all
agencies and departments that participate in the Network and
Information Technology Research and Development (NITRD) committee.
To date, I believe that the Department of Homeland Security's
Science and Technology Directorate has made excellent progress toward
meeting some of the goals outlined in the National Strategy to Secure
Cyberspace. We need to stay the course and bring these important
research and development products into the marketplace. But more needs
to be done if we are to counter the negative forces that threaten our
cyber security.
Based on the previously cited reports which reflect the views of
the professional community and in concert with our customers, the DHS
S&T Cyber Security program has identified the following research areas
as priorities which we will continue to address in FY2007, FY 2008 and
beyond:
We must continue to advance the development and
accelerate the deployment of more secure versions of
fundamental Internet protocols and architectures, including
those for the domain name system and routing protocols
described earlier.
We must improve and create new technologies for
detecting attacks or intrusions, including monitoring
technologies.
We must improve and create new methods for mitigation
and recovery, including techniques for containment of attacks
and development of resilient networks and systems that degrade
gracefully.
We must develop and support infrastructure and tools
to support cyber security research and development efforts,
including modeling and measurement, test beds, and data sets
for assessment of new cyber security technologies, such as the
DETER and PREDICT programs I described earlier.
We must assist the development and support of new
technologies to reduce vulnerabilities in process control
systems.
We must test, evaluate, and facilitate the transfer of
new technologies associated with the engineering of less
vulnerable software and securing the IT software development
lifecycle.
We need research to identify new solutions to address
malicious software, such as botnets and other ``malware,'' for
which no secure solutions currently exist.
We must develop trusted systems, new hardware and
software architectures for security, and develop cyber security
metrics.
We must develop tools that will allow us to visualize
network data so we can see where attacks are coming from and
diagnose cyber security problems faster and with more accuracy.
We must develop new ways to detect and mitigate
insider threats in cyber security.
We must develop the architecture and solutions that
will allow us to handle identity management on a wider scale
than is currently possible.
I want to stress for the Subcommittee that research and development
involves both promise and progress. The promise lies in our ability to
identify threats and potential solutions. But as long as these vital
research and development questions remain unanswered, they threaten all
of the progress we have made to date, creating weaknesses and
vulnerabilities that further complicate our task. The same is true for
the areas where we have already made valuable steps forward.
We need to deploy the important infrastructure protections we have
helped to develop--across the government and throughout the private
sector--and provide incentives for industry to partner in R&D efforts.
We need to move forward the already identified next-generation cyber
technology research projects that take aim at weaknesses we know today.
And we must continue to deliver tested technologies that can become
commercially available products, to extend the benefits of our research
and offer protection against cyber threats to homes and businesses
across the Nation.
The good news, Mr. Chairman and Members of the Subcommittee, is
that our research and development efforts show promise in addressing
the Nation's cyber security needs. I look forward to working with you
to advance our R&D efforts and address the security needs of our
Nation's critical infrastructure.
Mr. Langevin. I want to thank the panel for their
testimony.
I want to remind each member he or she will have 5 minutes
to question the panel, and I now recognize myself for
questions.
Dr. Saydjari, let me begin with you. You gave a pretty
sobering assessment which you laid out. I would like to ask the
panel to comment on what Dr. Saydjari testified to and if you
agree with the assessment. If not, would you expand on that?
Dr. Geer.
Mr. Geer. Well, sure. The threat is real. We have been, to
a large degree, lucky that we haven't seen it in grander form,
that there hasn't been a major episode.
One could say that--it is quite natural for most people--I
expect everybody in this room, certainly my family, for
example, to say that, because nothing big has happened that
they are aware of, that somehow the risk must not be as great
as people like Dr. Saydjari or myself or other members of the
panel say it is.
If you would accept the idea that if we have ever escaped a
bad event sheerly by luck, that at least you can put behind
yourself the argument that the absence of any major episode to
date is reassuring, I can give you one thought experiment that
illustrates that we have at least once avoided major disaster
by accident. It would be this.
9/11 riveted the country. Everybody paid attention, et
cetera. A week later a then-the-worst-we-had-ever-seen virus
came by, something called Nimda. Like most virus writers, the
person involved--like most good virus writers, amongst other
things this person left behind what is called a back door, an
ability to reenter the computer that they had previously
invaded, but by simpler means. So even if it turned out we knew
how Nimda got in in the first place and we closed that door,
there would be another door remaining behind. A little bit like
if I broke into your house and made a house key.
That idea of leaving behind a new back door is interesting.
Nimda at the time spread faster than we had ever seen anything
spread. Hands down the fastest we had ever seen. Since then,
there have been faster still, but at the time it was the
fastest we had ever seen.
Since all old viruses can be found somewhere on the
Internet at any given time, they never actually go away, let me
bring one of them up.
In 2001, a great deal of the Internet was still dial-up. A
lot of people accessed it by dial-up. There is a virus called
E911 which causes your modem to dial 911 constantly. When I
call you on the telephone, the line doesn't drop until I hang
up. When I call 911 on the telephone, the line doesn't drop
until you hang up, because you don't want the police to be able
to say who was I talking to when somebody cuts the wire.
Consequently, you can saturate a 911 console.
Where we got lucky, no clown had the bright idea to chase
the Nimda virus using its newly installed back door and install
the E911 virus cross-country. Because, if they had, all 911
services in the U.S. would have gone off the air in a matter of
a couple of hours. That would have had, if nothing else, been a
gran mal seizure of the public confidence.
So if you accept the argument that we have at least once
escaped a major event by dumb luck, then I think you can put
behind yourself any argument that is it really a big deal or
not. It really is a big deal if at least once we can show we
have escaped a major problem by dumb luck, and I think I just
gave you one.
Mr. Langevin. Thank you.
Dr. Lewis.
Mr. Lewis. Thank you.
I am the skunk at the party here because I don't really
agree with this. Part of the reason I don't agree with it is
because I do have some military experience, not as a member of
the military but as somebody who worked closely with them; and
I know how hard it is to derail a country. Even a third-world
country turns out to be much harder than we might suspect. Let
me tell you the reasons that I do think that.
While we do face serious problems on the informational
side, on the intelligence side, I think some of the other risks
are easy to overestimate. Some of the research that I would
base this on comes out of the strategy bombing survey that was
conducted at the end of the World War II by the United States,
and I would be happy to provide the committee with additional
information.
The first thing you have to ask yourself, though, is how
resilient is a country? If there is one attack, people don't
sit around; they respond. And so how long will it take people
to get back on line or to restore some kind of service?
The second thing you want to ask is, a very big country
turns out to be hard to derail; and you have all had this
experience. The experience I usually refer to is Charlotte,
North Carolina, which was taken off line for a week. No one
knows about it because of snowfall and they had electronic
power outrageous and all that.
You can remove major cities from the power grid and
telecommunications network. It has no effect on our military
power or, honestly, on our economy. A lot of this has to do
with political leadership and culture.
One of the things I have said in the past is, if we were
perhaps one of the more feeble European countries, if we were a
more excitable country, when there was an outbreak maybe we
would collapse. We have seen that happen. We have seen it
happen in the past. But I think Americans are a little tougher.
A lot depends on the leadership they see. If their leaders say
the right things, they will respond the right way.
Finally, you want to ask yourself how interconnected are
networks. There are few networks that are tightly
interconnected, whether it is electrical, telecom, the
financial network. These are things where you could have a
national level attack and you could have that kind of affect,
but most of the other stuff isn't that connected. So if you
knock out one city or one State or one water company, you are
not going to have a national effect.
So, for me, we need to look at the informational attacks,
we need to look at espionage, and we need to look at a few
critical networks that are interconnected. That is where there
is risk. I am a little more relaxed on some of the other
things.
Mr. Langevin. Dr. Maughan.
Mr. Maughan. I would have to say I agree with Mr. Saydjari
in his discussion.
I will remind you that our enemies are going to continue
advancing their capabilities and their technologies. And so
while we may decide to sit still and that we are OK they are
going to continue to advance and things are only going to get
worse. I believe the investment that he called for is at a bare
minimum to just keep up and may not even get us ahead.
Mr. Geer. May I add something, if I could, on this?
Mr. Langevin. Briefly, please.
Mr. Geer. This is a definitional question, perhaps, back to
you.
An attack that breaks things versus an attack that breaks
public confidence, what I spoke to was something that breaks
public confidence. I think the public confidence in, for
example, our financial networks can be broken without making
the entire network lay down and stay down. And so I guess
perhaps what we should be pushed about is define collapse or
define breakage. We may be in violent agreement once we get
past that.
Mr. Langevin. Dr. Saydjari, would you care to comment on
what you heard, particularly with Dr. Lewis' comments?
Mr. Saydjari. Yes, I would.
I think, first, I would point out there are 50 of the
Nation's leaders signing this letter of the President
estimating this risk at this level, including a former DCI, a
former director of NSA and a former director of DARPA. That is
no small level of talent in making this estimation.
The second thing I would point out is that we did a very
detailed analysis for this very reason, because there are
people who believe that the threat is overestimated. We took a
risk in developing this mock campaign against the United States
to develop it to prove that this is possible, and so we believe
that there is evidence that stands that says that this threat
is possible. Every part of that attack analysis was vetted with
various government agencies and the various sectors that were
involved in the attack, including power, including oil and gas,
financial service sectors and telecommunications.
We believe firmly in our analysis, and we believe that it
stands on its own merits, and we invite an independent
evaluation and an extension. That is indeed what we meant by
calling for a national threat assessment to validate our
findings and extend them so that we can develop sound policy
and settle this debate as to whether the threat is higher or
lower than what we are estimating.
Mr. Langevin. Thank you.
I am going to have other questions for the panel that you
may have to respond to in writing, but my time is expired so I
am going to yield to the gentleman from Texas, Mr. McCaul, for
5 minutes.
Mr. McCaul. I thank the Chair.
There is so much to talk about here I sometimes don't know
where to start. I am going to have to leave after this for a
briefing from General Petraeus. So if the Chair would indulge
me, I would like to throw everything out in one question.
Dr. Geer, you said what you can't see is more important
than what you can; and I agree with that. I think the threat of
the Trojan horse in this scenario is perhaps more devastating
than what we can see.
Dr. Lewis, you talked about foreign agents broke into the
Department of Defense and stole file cabinets. That would cause
hysteria in the media. And yet we know we have intrusions in
the Federal Government's networks, and I don't know if we have
an idea as to what is being stolen.
You talked about metrics. I don't think we can gauge or
hold accountable if we don't know what they are taking. A
technical idea of attack data. We don't know where these
attacks are coming from, but we know they are coming. And there
are several levels of these attacks. One may be purely for
mischief, one could be criminal, another espionage. As you
point out, I think we talked a little bit about China and its
willingness and capacity to steal information, steal secrets,
intellectual property theft.
But the last scenario that Mr. Saydjari really kind of
focuses on is one that really keeps me up at night, and that is
the idea of a cyber attack that is along the lines of warfare.
An attack we know that our own military is capable of doing and
shutting down power grids in other countries, yet we don't know
what some of these rogue nations, what their capacity and
capability really is. We do know that any nation with a power
grid can probably figure out how to shut it down.
I think the ramifications are--I know, Dr. Lewis, maybe
there has been some exaggeration, but maybe not. To the extent
this country could be shut down, albeit temporarily, I think
the destruction it would cause is very clear.
The idea of a national threat assessment just to gauge
where are we, you threw a number out that is about 10 times
more than what we authorize in R&D for cybersecurity. And I
throw this out to the panel, and I appreciate the Chair
indulging me on the time, but if you could talk, first of all,
Mr. Saydjari, about the threat assessment that you did and what
the results were and then possibly talk about--when you say
vulnerable to nation-state adversaries, who do you think they
are, specifically? And then I will open it up to the panel for
just a full discussion.
Mr. Saydjari. Sure. The development that we do is called
Dark Angel. This was a mock attack by seven of the leaders,
Professionals for Cyber Defense, this nonprofit group; and we
developed a detailed attack tree against our Nation. The
purpose of it was to do a strategic blow to our country; and we
looked at various domains, including the financial services
sector, telecommunications, power, oil and gas. We looked at
all of them.
One of the things that has been lacking to date is sort of
an isolated look at each of the domains. What we looked at is
looking at it from a nation-state's perspective about doing
strategic damage and looking at the interconnections between
those domains and doing a campaign, including rolling attacks
on various symptoms. Once they recover, attack them again.
Attack in a way that actually disables physical things, like
power generators.
We are not talking about small-scale power outages for a
day or two. We are talking about destroying power generators by
improper control. We are talking about blowing up transformers
by improper control. And these generators and transformers take
months to remanufacture. And, oh, by the way, some of them we
can't manufacture in the United States anymore. We have to go
to Europe to get it. So if that attack happens in Europe at the
same time, guess who is going to get priority on those
transformers and power generators?
So we did this detailed analysis. We have this very, very
sophisticated attack tree that has been deeply vetted by
various domain experts. We did this over the course of 30 days
in response to a comment on the President's national strategy
to sort of put up our position that there was a serious
national threat and we were forced into developing this
scenario. And we believe it is absolutely compelling.
Again, we don't make this publicly available, but we invite
a limited review to say, OK, you don't think the threat is this
bad? Great, come look at what we did, extend what we did.
Mr. Langevin. Will the gentleman yield for one second?
You said this is in the context of an attack from a nation-
state. Could it also translate over into a rogue individual or
individuals such as a terrorist group carrying out the same
level of attack with the same type of catastrophic
consequences?
Mr. Saydjari. Our assumption was a $500 million budget and
about 3 years of preparation. So an individual certainly could
not do this. But a transnational terrorist group like al-Qa'ida
certainly could. In fact, that was our model as a transnational
terrorist organization or a small nation-state. Certainly a
large nation-state is well within their means and well within
their patience.
And I point out also that we are not just assuming cyber
attacks, we are assuming insider attacks, we are assuming
malicious code, we are assuming lifecycle attacks, where
somebody attacks the code that is being developed and gets code
that blows up on us on the fly at their discretion. So we are
talking about a very sophisticated attack from a military
perspective against the United States.
Mr. McCaul. Again, that is my greatest fear, particularly
if it comes from a terrorist rogue nation. Did you brief the
Department of Homeland Security on this assessment?
Mr. Saydjari. Yes, sir. In about the March or April time
frame of 2005 we did do that briefing. And they politely heard
our briefing, and we saw no follow-up activity or actions from
that briefing.
Mr. McCaul. Is that correct, Dr. Maughan? Was there no
response?
Mr. Maughan. That briefing was provided to the National
Cybersecurity Division, not to the Science and Technology
Directorate.
Mr. McCaul. And so you can't answer on behalf of anything
outside your Directorate?
Mr. Maughan. Correct.
Mr. McCaul. Do you think the idea of a national threat
assessment is a good idea?
Mr. Maughan. Yeah. The Department has been out doing
physical assessments of a lot of the critical infrastructure
owned and operated by the private sector. We should do a
similar from a cyber perspective, both government and industry,
given that industry owns and operates a significant portion of
that infrastructure.
Mr. McCaul. I personally think it would be a good idea to
be able to measure that, as Dr. Geer talks about, the metrics.
Can you comment about this kind of worst-case scenario?
Mr. Geer. Sure. You mean, give you an example of one?
Mr. McCaul. Yes.
Mr. Geer. Do you want to take the Internet down this
afternoon?
Mr. McCaul. I kind of would like to stay out of jail.
Mr. Geer. Well, so would I. Figure out how to worm IOS,
which is the operating system for Cisco routers, which dominate
the top level of the Internet. Go in and have them rewrite the
EPROMs as fast as you can go. 50,000 cycles, they burn out, you
now have to have to visit it with a soldering iron 3 minutes.
Mr. McCaul. Dr. Maughan, do you consult with experts like
Dr. Geer in terms of anticipating vulnerabilities?
Mr. Maughan. We do, and we try to bring in the experts
every chance we can.
Mr. McCaul. I would highly recommend it.
Dr. Lewis, any comment.
Mr. Lewis. I want to take the contrary view again. Some of
us call these weapons of mass annoyance. If we are talking in
military terms, let's talk in military terms. I am China and I
go to make your traffic lights blink on and off for a week or
so. Is that going to stop the carrier battle groups from going
to the Taiwan Straits? Is it going to reduce American military
capabilities? Is it going to damage the American economy over
the long term? The answer is no.
So if you are a Chinese leader, you think I am going to do
something, it is going to really irritate them, they will be
mad, and I am not going to get any military benefit from it.
And that is how I think about it.
Now a rogue state, perhaps their calculus will be a little
different. It is hard to predict when they are so crazy like in
North Korea or Iran. A terrorist group probably doesn't have
the capabilities.
But when you look at the people who are likely to do this,
they are asking themselves, what do I get out of it? How likely
is it to make me better off in a conflict? And, right now, they
don't think it is going to make them better off.
Mr. McCaul. And I agree with you. China is all about
espionage and intellectual property. But there are other
organizations out there. And when teenagers can hack into
computers, it is a little disturbing to think of the
destruction that could be caused by someone who has this
ability, someone who has it in the wrong hands. And I think
when we know the terrorist's main goal is to destroy
preliminarily our financial markets, it raises the bar.
That is really all I have, Mr. Chairman, but I want to
thank all the witnesses for being here today. It has been very
insightful. Thank you.
Mr. Langevin. The gentleman from Texas, Mr. Green, is
recognized for 5 minutes.
Mr. Green. Thank you, Mr. Chairman; and thank you for
hosting these important meetings and hearings.
My question has to do with punishment. What has been your
experience in terms of persons who are caught? How are they
punished?
Someone gave the example of someone breaking into an
administrative office and taking files. My suspicion is we call
that a felony and the person would be severely punished. What
is your experience with reference to cyber theft?
Mr. Lewis. I have done a little research on that, and my
experience and what I have learned from the FBI and from other
law enforcement agencies is you are not going to be caught, and
it is almost a risk-free crime. We don't have a good metric. It
is true. So is it 95 percent of the people who do this escape?
Is it closer to 100 percent? Is it a bit less? But the odds
are, if you engage in a cyber attack, if you steal information,
if you break into someone's network, particularly if you do it
from overseas, it is a risk-free event.
Mr. Green. Any other opinions? Everybody is in agreement
that it is risk free?
What about encoding? Is that something that we can hope to
have some sort of safety with, some sort of encryptions for
specific areas of security concerns?
Mr. Geer. I can say that, in the commercial sector,
adoption of encryption at one level or another is going about
as fast as it can go. That is not to say it is slow. They are
spending money like crazy to encrypt.
The common thing that appears in the newspaper is I lost
the laptop in the cab kind of thing or somebody broke into my
house. That kind of thing is going as fast as it can go. I
think that you probably will see in a matter of years nearly
nothing that isn't encrypted, where the general counsel is
aware that the company has it.
Beyond that, do you want all transactions and so forth, all
communications over the net to be encrypted? Maybe. It is not a
be-all and an end-all. It helps.
I think that you should remember that encryption is,
generally speaking, no solution to the insider problem. So you
might be able to get rid of a degree of the outsider problem,
but you would not get rid of the insider problem by adopting
full tilt encryption ideas.
Mr. Saydjari. I would like to add to that.
So encryption is a very valuable tool, particularly
protecting information in transit. But one of our biggest
problems is the security at the host. And, ultimately, the data
has to be decrypted at the end machines to actually do
something with it; and these are the places that we are most
vulnerable. So I am a very strong advocate of getting
encryption out there in a widespread way and making it
available to the private sector and having it proliferate, and
it will help. But I just want to make sure that we all
understand that processing at the host and things like denial-
of-service attacks on the availability of those hosts are
affected in no way by encryption.
Mr. Maughan. I would agree with what Mr. Saydjari has said.
Cryptography is only going to do a small amount for us in a big
picture. There are bigger problems to our end system's
vulnerability. Encryption is only one tool in the quiver of
arrows that we have.
Mr. Green. Is it fair to say that we may never be able to
become completely secure because as we get better it seems that
there is always a new thought or hype, idea, in terms of making
the invulnerable vulnerable?
A comment please. I like your smile, Dr. Geer. Let me hear
your comment.
Mr. Geer. No. Perfection is impossible because it involves
dividing by zero and you can't afford the cost. This is purely
a risk management problem.
If you really want my car, you can probably get it. I can
lock it in the garage, I can lock the car, et cetera, et
cetera. The guy with a blowtorch and a tow truck and a heavy
lift helicopter can still probably get it. I can, however, make
my neighbor's car a lot more attractive than mine; and to a
degree that is all that we can do here. All we can do is make
it such that the people who want stuff that we do not want them
to have, have to go somewhere else.
Mr. Geer. And I know that sounds unfortunate, but I think
that is the right mindset to have. Maybe you will have a happy
surprise, and you do actually solve a problem on getting rid of
smallpox or polio or something, but generally speaking, you
cannot get rid of it. What you can do is make it harder. You
can make them go somewhere else.
Mr. Green. I see other smiles, so let us go with the next
smiling face.
Mr. Saydjari. I completely agree with Dan. I think the
threat is always going to be escalating. There is always going
to be higher degrees of integration of our systems and new
capabilities in our systems that will be attackable, and one
thing, I think, we all have to understand here is that this is
not a one-shot investment. So, when I talk about a multi-
billion dollar program establishing a capability in 3 years, it
is not done in 3 years. It is a sustaining investment to be
actively engaged in the escalation that will inevitably happen
as we have seen over the last 10 years. The level of
sophistication of attacks has risen dramatically over the last
10 years. The kinds of attacks we have seen in the wild are
amazingly complex and amazingly sophisticated, and we will only
see them get worse in terms of the level of damage they do.
Mr. Maughan. I was only going to agree with them.
It is a cat-and-mouse game that we are playing with the bad
guys, and we are never going to be able to secure our systems
100 percent, and so the best we can do, as Dr. Geer said, is
risk management and try to defend our systems as best we can.
Mr. Lewis. We are all in tremendous agreement here, but I
want to put a little different cast on it, which is let us not
think defensively. We cannot make them perfectly secure, but we
just want to be in a position where we do better than our
opponents. So that is a good goal. If we get more out of this
than our opponents do, we win.
Mr. Green. Thank you, Mr. Chairman. I yield back.
Mr. Langevin. Thank you.
The gentleman from North Carolina, Mr. Etheridge, is
recognized for 5 minutes.
Mr. Etheridge. Thank you, Mr. Chairman, and let me thank
you for holding this hearing.
Gentlemen, what we are seeing in the 21st century is going
to be a huge challenge. Last week, before this committee, we
heard from Federal agencies--Commerce and State--talking about
the attacks on their systems that were unexpected, but they may
not have even been aware of them until well after they had
occurred, and further, even after the illegal access was
noticed, the date and duration of attacks could not be
determined, and the extent of information compromised may never
be known. That is what they shared with us. So my question to
you is:
Is it ever possible to determine after an attack the extent
of the damage? You know, for example, can logs be altered or
so-called rogue tunnels be constructed to hide the nature of
the attack? Do you agree? The answer is ``yes''? Everybody
agrees. OK.
So my next question is: What tools do we have available to
us to identify the attacks, which seem to me to be critical,
and to check the authenticity of the date so that we know when
the attacks occurred, and to the extent we know that, how can
we deal with it?
Who wants to tackle that first?
Mr. Geer. One of the hardest questions for most of us in
the commercial sector is: If you know something is going on,
how do you pursue it? Because it is a very fine line between
noticing it and then somehow finding yourself engaged in a
countermeasure. You know, do I have the right to--I was at a
workshop 2 weeks ago, and there are a couple of other people in
the room who were at this same workshop. If I discover what is
called a ``robot network''--or a botnet--in my firm, if I
discover that in my firm someone has taken over a number of
computers and they are being used for purposes nefarious, do I
have a right to disable that botnet? Do I have a right to
poison the command and control system that it uses to operate?
Do I have a right to take them off the air from where I sit?
Now, at the moment, that is, I think, roughly equivalent
to, ``well, if nobody knows, your general counsel would advise
you not to,'' but in this space, there is a very fine line
between how do you defend yourself and what somebody else will
later charge as vigilantism.
Mr. Etheridge. Let me interrupt you if I might.
If, prior to the computer, my files were in file cabinets
and you come and lift out those files and, in fact, take them
with you, you are in trouble.
Mr. Geer. Yes.
Mr. Etheridge. This is the same kind of thing except you
are doing it electronically from a remote site which may be two
times removed.
Mr. Geer. Yes, but the difference there is, if I steal your
car or your files, you know they are gone. If I steal your
data, you may not know it is gone until it is misused. So I
have to be able to react when I discover that it is going on.
Whether this is ``the home is the castle, and I can shoot the
intruder or not,'' I mean, I do not know quite what to say
here, but this is a problem. This is the fundamental problem on
the commercial side.
Mr. Etheridge. Please. We are looking for some R&D, some
way we can get there because this, to me, seems to be that key
we have got to find to either lock the lock or unlock the lock
that we have got to get to.
Mr. Saydjari. So I think this is partly a question of
intrusion detection systems, and the intrusion detection
systems that are out there today really count on the attack's
having been seen in the wild before. They are called
``signature-based schemes,'' and they are ineffective in the
sense that they are after the fact, and so a majority or
certainly a very large number of attacks that are out there are
not visible by these kinds of mechanisms, and that is a bad
thing, and there is research, for example, on anomaly-based
detection schemes that can characterize normal behavior and
then look for the abnormal behavior, which is a deviation for
that. So there is hope on that research line.
I will also add that the community has been using what I
would consider ad hoc sensors, sorts of things that were not
really designed to be sensors for the most sophisticated kinds
of attacks like the ones that we imagine and work through in
the dark angel campaign. So what we really need to do as a
community is to work backwards from the kinds of attacks we are
most worried about to the kinds of sensors that we require to
detect those. I mean it is like, you know, if we were trying to
detect a nuclear launch just to kind of look for, you know,
some warm sensations from somebody nearby. I mean we cannot
just use those kinds of off-the-shelf kinds of sensors. We
really need to rethink the way we do sensors.
Mr. Lewis. Let me offer you a suggestion that is maybe a
little less expensive and will not cost as much money.
One of the problems that I think we have seen is sometimes
there is knowledge in the national security communities and the
national security agencies like defense or the intelligence
community that does not get shared or does not get shared
promptly or adequately with the civilian agencies. That might
be an interesting thing for you to look at. So, if DOD figures
out there is a problem, how does that percolate through the
rest of the Federal system?
Mr. Etheridge. How do we get out of the tunnels and start
sharing at the highest level?
Mr. Lewis. Exactly. So better coordination, better
information--sharing, breaking that firewall between, say, some
of the national security folks. That would help.
The other thing that would help would be better network
hygiene for lack of a better term. Now, that will not solve the
problem, but it will reduce the number of incidents, and what
you have got is some network administrators do a great job;
other network administrators do not do as good a job. How do
you get them all up to the a basic level? We have seen some
cases where, at NASA or at DOD, grabbing the low-hanging fruit
has significantly reduced the number of incidents. The systems
are not secure. People are still intruding, but it is at a much
lower level.
Mr. Saydjari. If I could extend my remarks at one more
level, a colleague of mine who is an expert in the power system
advises me that, if we had an attack on our power control
systems, we would never know it because there are no intrusion
detection systems within those networks. So, when Dr. Lewis
talks about the focus on the networks that are connected, I
will tell you that every network is connected to every other
network in some way, shape or fashion, whether it is through
software development or actual connections, and so those
networks are just as likely to be attacked. Well, of course,
you need some insiders or you need some malicious software, but
you can attack those networks, and those networks which are
controlling our most critical assets are least sensored. That
is a very bad thing that needs to change immediately.
Mr. Geer. I like numbers. Can I give you a couple?
Mr. Etheridge. Please.
Mr. Geer. For average desktop machines--I am not talking
about, for example, the power grid. For average desktop
machines, my own calculation is that about 30 percent of them
have something unwanted running on them. Vent Surf says 40
percent; Microsoft says two-thirds; IDC says three-quarters. So
it is not like we are trying to preserve innocence. It is a
little harder.
Mr. Etheridge. Yes. Well, you have scared me to death.
Thank you.
Mr. Langevin. Well, gentlemen, I want to thank you for your
testimony today. You, obviously, addressed and raised some very
sobering and very serious issues, and we obviously have a lot
of work to do. We look forward to speaking with you further.
I am sure that other members of the committee, myself
included, will have additional questions that we might want to
pose to you, and we ask that you respond, if you would, in an
expeditious manner. If you could help us with that, we would be
very grateful and would much appreciate it. Thank you very
much, and I want to thank the witnesses for their testimony.
Hearing no further business before the subcommittee, the
subcommittee is adjourned.
[Whereupon, at 2:20 p.m., the subcommittee was adjourned.]
Appendix I: For the Record
----------
Prepared Statement of the Honorable Bennie G. Thompson, Chairman,
Committee on Homeland Security
I thank the Chairman for holding another important hearing
on cybersecurity.
It is clear that our government, working together with the
private sector and academia, must do more to ensure that cybersecurity
is a priority in our nation's homeland security strategy.
In 1996, the United States government undertook the first
national effort to secure our networks.
Unfortunately, I don't believe that we are any further
along today in our efforts to secure cyberspace.
Programs and initiatives that were developed over the past
ten years have been dismantled and, in certain instances, are just now
being re-created by the government.
We heard in last week's hearing that ``coordinating better
cyber security practices across the Federal government'' is one of
Secretary Chertoff's ``highest priorities.''
But this rings hollow to me when I think about how long it
took him to appoint an Assistant Secretary for Cybersecurity.
I also wonder why the Secretary believes that the
Department will be able to coordinate better cyber security practices
across the Federal government, when his own Chief Information Officer
just received a ``D'' in the recent FISMA grades.
So we have a lot of work to do, but fortunately we have
some very capable people who can help.
I thank the witnesses for being here today and for their
commitment to helping the Federal government move this issue in the
right direction.
Thank you Mr. Chairman.
Appendix II: Selected Major Reports on Cyber Security Research and
Development
----------
Biometric Research Agenda: Report of the NSF Workshop. Morgantown, West
Virginia, April/May 2003, http://64.233.167.104/
search?q=cache:xweu9dx2qMsJ:www.
wvu.edu/bknc/
BiometricResearchAgenda.pdf+Biometric+Research+Agenda:+Report+
of+the+NSF+Workshop&hl=en&ct=clnk&cd=3&gl=us.
Coordination of Federal Cyber Security Research and Development, U.S
Government Accountability Office, GAO-06-811, Sept. 2006, http://
www.gao.gov/new.items/d06811.pdf.
Creating a National Framework for Cybersecurity: An Analysis of Issues
and Options, Eric A. Fischer, Congressional Research Service, Feb. 22,
2005, http://www.au.af.mil/au/awc/awcgate/crs/rl32777.pdf.
Critical Foundations: Protecting America's Infrastructures. President's
Commission on Critical Infrastructure Protection, October 1997,
www.fas.org/sgp/library/pccip.pdf.
Critical Information Infrastructure Protection and the Law: An Overview
of Key Issues. Computer Science and Telecommunications Board, National
Research Council, 2003, http://www.cstb.org/pub_ciip.html.
Critical Infrastructure: Challenges Remain in Protecting Key Sectors,
Testimony of Eileen R. Larence, Director, Homeland Security and Justice
Issues, and David A. Powner, Director, Information Technology
Management Issues, Before the Subcommittee on Homeland Security,
Committee on Appropriations, House of Representatives, U.S. Government
Accountability Office, GAO-07-626T, March 20, 2007, http://www.gao.gov/
new.items/d07626t.pdf.
Critical Infrastructure Protection: Challenges and Efforts to Secure
Control Systems, Testimony of Robert F. Dacey, Director, Information
Security Issues, Before the Subcommittee on Technology Information
Policy, Intergovernmental Relations and the Census, House Committee on
Government Reform, U.S. Government Accountability Office, GAO-04-628T,
March 30, 2004, http://www.gao.gov/new.items/d04628t.pdf.
Critical Infrastructure Protection: Challenges in Addressing
Cybersecurity, Testimony of David A. Powner, Director Information
Technology Management Issues, Before the Subcommittee on Federal
Financial Management, Government Information, and International
Security, Senate Committee on Homeland Security and Governmental
Affairs, U.S. Government Accountability Office, GAO-05-827T, July 19,
2005, http://www.gao.gov/new.items/d05827t.pdf.
Cyber Security Research and Development Agenda. I3P, Dartmouth College,
January 2003, http://www.thei3p.org/repository/
2003_Cyber_Security_RD_Agenda.pdf.
Electronic Crime Needs Assessment for State and Local Law Enforcement,
National Institute of Justice Research Report, March 2001, http://
www.ncjrs.org/pdffiles1/nij/186276.pdf.
Embedded, Everywhere: A Research Agenda for Networked Systems of
Embedded Computers. Computer Science and Telecommunications
Board,National Research Council, 2001, http://
www7.nationalacademies.org/cstb/pub_embedded.html.
Hard Problems List. Infosec Research Council. September 1999 (and draft
revision as of September 2004) Information Technology Research for
Crisis Management. Computer Science and Telecommunications Board,
National Research Council, 1999, http://www7.nationalacademies.org/
cstb/pub_crisismanagement.html.
High Confidence Software and Systems Research Needs. High Confidence
Software and Systems Coordinating Group, Interagency Working Group
onInformation Technology Research and Development, January 2001, http:/
/www.nitrd.gov/pubs/hcss-research.pdf.
IDs-Not That Easy. Questions About Nationwide Identity Systems.
Computer Science and Telecommunications Board, National Research
Council, 2002, http://www7.nationalacademies.org/cstb/
pub_nationwideidentity.html.
Information Sharing/Critical Infrastructure Protection Task Force
Report, National Security Telecommunications Advisory Committee, May
2000,
http://www.ncs.gov/nstac/reports/2000/ISCIP-Final.pdf.
Information Technology for Counterterrorism. Computer Science
andTelecommunications Board, National Research Council, 2003, http://
www7.nationalacademies.org/cstb/pub_counterterrorism.html.
Insider Threat Study: Computer System Sabotage in Critical
Infrastructure Sectors, Michelle Keeney, Dawn Cappelli, et al, Carnegie
Mellon Software Engineering Institute, May 2005, http://www.cert.org/
cert/work/organizational_security.html.
Internet Domain Names: Background and Policy Issues, Lennard G. Kruger,
Congressional Research Service, Sept. 22, 2005, http://www.au.af.mil/
au/awc/awcgate/crs/97-868.pdf.
The Internet Under Crisis Conditions: Learning from September 11.
Computer Science and Telecommunications Board, National
ResearchCouncil, 2003, http://www7.nationalacademies.org/cstb/
pub_internet911.html.
National Security Telecommunications Advisory Committee, Research and
Development Exchange Workshop, Atlanta, Georgia, March 2003, http://
www.ncs.gov/nstac/rd/nstac_03_bos.html.
National Security Telecommunications Advisory Committee, Research
andDevelopment Exchange Workshop. Tulsa, Oklahoma, September 2000,
http://www.ncs.gov/nstac/reports/2001/R&D_Exchange2000Proceedings.htm.
National Security Telecommunications Advisory Committee, Research and
Development Exchange Workshop. West Lafayette, Indiana, October 1998,
http://www.ncs.gov/nstac/reports/1998/R&DExchange.pdf.
National Strategy to Secure Cyberspace, The White House, February 2003,
http://www.whitehouse.gov/pcipb/.
Protecting Systems Task Force Report on Enhancing the Nation's Security
Efforts, National Security Telecommunications Advisory Committee, May
2000,
http://64.233.167.104/search?q=cache:JkJUKZ9OmYsJ:www.ncs.gov/nstac/
reports/2000/PSTF-Final.pdf+Protecting+Systems+Task+Force+Report+on+
Enhancing+the+Nation%E2%80%99s+Security+Efforts,+National+Security+
Telecommunications+Advisory+Committee,+May+2000,&hl=en&ct=clnk&
cd=1&gl=us.
Robust Cyber Defense. Study commissioned for DARPA ITO, Fall 2001.
Slides available at:
http://www.cs.cornell.edu/fbs/darpa.RobustCyberDefense.ppt.
Technology Assessment: Cybersecurity for Critical Infrastructure
Protection, U.S. Government Accountability Office, GAO-04-321, May
2004, http://www.gao.gov/new.items/d04321.pdf.
Trust in Cyberspace. Computer Science and Telecommunications
Board,National Research Council, 1999, http://books.nap.edu/
readingroom/books/trust/.
Understanding the Insider Threat, Richard C. Brackney, Robert H.
Anderson, Conference Proceedings of a March 2004 Workshop, RAND,
National Security Division, http://www.rand.org/pubs/conf_proceedings
/2005/RAND_CF196.pdf.
Who Goes There? Authentication Through the Lens of Privacy. Computer
Science and Telecommunications Board, National Research Council, 2003,
http://www7.nationalacademies.org/cstb/pub_authentication.html.
Workshop on Scalable Cyber-Security Challenges in Large-Scale Networks:
Deployment Obstacles. Large Scale Networking Coordinating Group,NITRD,
Landsdowne, Virginia, March 2003, http://64.233.167.104/
search?q=cache:mWKvtoq--xLoJ:cs.
yale.edu/homes/jf/LSN-report.pdf+Workshop+on+Scalable+Cyber-Security+
Challenges+in+Large-Scale+Networks:&hl=en&ct=clnk&cd=1&gl=us.