[House Hearing, 108 Congress]
[From the U.S. Government Publishing Office]




CAN THE USE OF FACTUAL DATA ANALYSIS STRENGTHEN NATIONAL SECURITY? PART 
                                  ONE

=======================================================================

                                HEARING

                               before the

                SUBCOMMITTEE ON TECHNOLOGY, INFORMATION
                POLICY, INTERGOVERNMENTAL RELATIONS AND
                               THE CENSUS

                                 of the

                              COMMITTEE ON
                           GOVERNMENT REFORM

                        HOUSE OF REPRESENTATIVES

                      ONE HUNDRED EIGHTH CONGRESS

                             FIRST SESSION

                               __________

                              MAY 6, 2003

                               __________

                           Serial No. 108-72

                               __________

       Printed for the use of the Committee on Government Reform


  Available via the World Wide Web: http://www.gpo.gov/congress/house
                      http://www.house.gov/reform


                                 ______

90-399              U.S. GOVERNMENT PRINTING OFFICE
                            WASHINGTON : 2003
____________________________________________________________________________
For Sale by the Superintendent of Documents, U.S. Government Printing Office
Internet: bookstore.gpo.gov  Phone: toll free (866) 512-1800; (202) 512ï¿½091800  
Fax: (202) 512ï¿½092250 Mail: Stop SSOP, Washington, DC 20402ï¿½090001

                     COMMITTEE ON GOVERNMENT REFORM

                     TOM DAVIS, Virginia, Chairman
DAN BURTON, Indiana                  HENRY A. WAXMAN, California
CHRISTOPHER SHAYS, Connecticut       TOM LANTOS, California
ILEANA ROS-LEHTINEN, Florida         MAJOR R. OWENS, New York
JOHN M. McHUGH, New York             EDOLPHUS TOWNS, New York
JOHN L. MICA, Florida                PAUL E. KANJORSKI, Pennsylvania
MARK E. SOUDER, Indiana              CAROLYN B. MALONEY, New York
STEVEN C. LaTOURETTE, Ohio           ELIJAH E. CUMMINGS, Maryland
DOUG OSE, California                 DENNIS J. KUCINICH, Ohio
RON LEWIS, Kentucky                  DANNY K. DAVIS, Illinois
JO ANN DAVIS, Virginia               JOHN F. TIERNEY, Massachusetts
TODD RUSSELL PLATTS, Pennsylvania    WM. LACY CLAY, Missouri
CHRIS CANNON, Utah                   DIANE E. WATSON, California
ADAM H. PUTNAM, Florida              STEPHEN F. LYNCH, Massachusetts
EDWARD L. SCHROCK, Virginia          CHRIS VAN HOLLEN, Maryland
JOHN J. DUNCAN, Jr., Tennessee       LINDA T. SANCHEZ, California
JOHN SULLIVAN, Oklahoma              C.A. ``DUTCH'' RUPPERSBERGER, 
NATHAN DEAL, Georgia                     Maryland
CANDICE S. MILLER, Michigan          ELEANOR HOLMES NORTON, District of 
TIM MURPHY, Pennsylvania                 Columbia
MICHAEL R. TURNER, Ohio              JIM COOPER, Tennessee
JOHN R. CARTER, Texas                CHRIS BELL, Texas
WILLIAM J. JANKLOW, South Dakota                 ------
MARSHA BLACKBURN, Tennessee          BERNARD SANDERS, Vermont 
                                         (Independent)

                       Peter Sirh, Staff Director
                 Melissa Wojciak, Deputy Staff Director
                      Rob Borden, Parliamentarian
                       Teresa Austin, Chief Clerk
              Philip M. Schiliro, Minority Staff Director

   Subcommittee on Technology, Information Policy, Intergovernmental 
                        Relations and the Census

                   ADAM H. PUTNAM, Florida, Chairman
CANDICE S. MILLER, Michigan          WM. LACY CLAY, Missouri
DOUG OSE, California                 DIANE E. WATSON, California
TIM MURPHY, Pennsylvania             STEPHEN F. LYNCH, Massachusetts
MICHAEL R. TURNER, Ohio

                               Ex Officio

TOM DAVIS, Virginia                  HENRY A. WAXMAN, California
                        Bob Dix, Staff Director
                 Scott Klein, Professional Staff Member
                      Ursula Wojciechowski, Clerk
           David McMillen, Minority Professional Staff Member



                            C O N T E N T S

                              ----------                              
                                                                   Page
Hearing held on May 6, 2003......................................     1
Statement of:
    Loy, Admiral James L., Director, Transportation Security 
      Administration.............................................    31
    McCraw, Steve, Assistant Director, Office of Intelligence, 
      Federal Bureau of Investigation, accompanied by William 
      Hooten, Deputy Executive Assistant Director................    13
    Tether, Tony, Director, Defense Advance Research Project 
      Agency, Department of Defense..............................    46
Letters, statements, etc., submitted for the record by:
    Clay, Hon. Wm. Lacy, a Representative in Congress from the 
      State of Missouri, prepared statement of...................     8
    Loy, Admiral James L., Director, Transportation Security 
      Administration, prepared statement of......................    34
    McCraw, Steve, Assistant Director, Office of Intelligence, 
      Federal Bureau of Investigation, prepared statement of.....    16
    Miller, Hon. Candice, a Representative in Congress from the 
      State of Michigan, prepared statement of...................    25
    Putnam, Hon. Adam H., a Representative in Congress from the 
      State of Florida, prepared statement of....................     4
    Tether, Tony, Director, Defense Advance Research Project 
      Agency, Department of Defense, prepared statement of.......    49

 
CAN THE USE OF FACTUAL DATA ANALYSIS STRENGTHEN NATIONAL SECURITY? PART 
                                  ONE

                              ----------                              


                          TUESDAY, MAY 6, 2003

                  House of Representatives,
   Subcommittee on Technology, Information Policy, 
        Intergovernmental Relations and the Census,
                            Committee on Government Reform,
                                                    Washington, DC.
    The subcommittee met, pursuant to notice, at 3 p.m., in 
room 2154, Rayburn House Office Building, Hon. Adam Putnam 
(chairman of the subcommittee) presiding.
    Present: Representatives Putnam, Miller and Clay.
    Staff present: Bob Dix, staff director; John Hambel, senior 
counsel; Scott Klein, Chip Walker, Lori Martin, and Casey 
Welch, professional staff members; Ursula Wojchechowski, clerk; 
Suzanne Lightman, fellow; David McMillen, minority professional 
staff member; and Jean Gosa, minority assistant clerk.
    Mr. Putnam. A quorum being present. The hearing of the 
Subcommittee on Technology, Information Policy, 
Intergovernmental Relations and the Census will come to order. 
Good afternoon and welcome to today's hearing, ``Can the Use of 
Factual Data Analysis Strengthen National Security? Part One.'' 
First of all, I'd like to thank everyone for bearing with us as 
we've had to make time and room changes for today's hearing. We 
appreciate your cooperation.
    In an effort to prevent future terrorist attacks and 
enhance law enforcement efforts, deputies and agencies 
throughout the Federal Government have begun developing 
strategies that will assist in the identification of potential 
risks through the use of technology and information sharing. 
The truth is that there is a tremendous amount of information 
that already resides in the public venue. However, due to past 
practices of stovepipe mentalities and turf issues, much 
relevant information that could be of use or interest to law 
enforcement officials has not been easily accessible.
    In particular, since September 11, 2001, it has been 
imminently clear that we must do a better job of compiling and 
sharing information that will provide, enhance the 
opportunities for law enforcement and national security 
officials to identify potential risks in advance. Federal 
agencies have utilized methodologies that facilitate data base 
exploration for quite some time in an effort to root out waste, 
fraud and abuse. In fact the recent highly public case of 
government credit card abuse was flushed out, and the 
perpetrators identified through the use of data mining or 
factual data analysis, as some call it.
    Now, a number of Federal agencies with the responsibility 
for homeland security and law enforcement are employing the 
lessons learned through the use of factual data analysis or the 
conclusions drawn from this analytical process to increase 
their ability to detect patterns and relationships within the 
masses of data they have access to in an effort to increase 
risk assessment capabilities. This hearing will examine whether 
the use of this process will successfully enhance efforts to 
strengthen law enforcement and national security.
    Does factual data analysis contribute to increase the risk 
detection? As we have previously established, factual data 
analysis is not a technology in and of itself. It is an 
analytical process that utilizes technology in an effort to 
identify patterns and relationships that were previously 
unknown. It has been used successfully in the private sector to 
craft specific marketing and sales programs. It has been used 
successfully in the public sector to identify and address 
instances of waste, fraud and abuse.
    The hope is that these same technological advances that aid 
marketers in identifying customers for their products and law 
enforcement in catching tax evaders or identifying welfare 
fraud will also detect patterns that should raise suspicion 
among those working would improve our Nation's security. Today 
we have witnesses representing the FBI, the Transportation 
Security Administration, and Defense Advanced Research Projects 
Agency.
    Each of these three agencies proposes to use factual data 
analysis or conclusions drawn from the process to enhance 
homeland security. Specifically, we will be examining the FBI's 
Trilogy and related technology analysis tools, TSA's computer-
assisted prescreening process system [CAPPS] II, and DARPA's 
total information awareness [TIA]. We have asked each of these 
witnesses to explain their agency's program and talk about the 
role factual data analysis is envisioned to play. While each of 
these agencies proposals is different in its construct and each 
may generate varying responses and levels of interest, the 
subcommittee will seek to learn more about the source, 
accuracy, reliability, and security of the data that is 
accessed to determine risk assessment.
    Let me be clear, we are not here to compare one project to 
the other, nor are we here to evaluate the strategic basis for 
these projects. We are here to examine the use of technology in 
the facilitation of this process and the techniques, processes 
and outcomes that are produced. We hope to listen and learn 
from these expert witnesses and hear factual information about 
these projects. We also recognize that there is clearly some 
concern and reluctance on the part of some of the witnesses to 
even be here today because of some of the press coverage about 
these projects. Today we will examine the facts about how data 
will be compiled, what data will be assembled, what steps are 
taken to ensure the accuracy and reliability of the data, how 
the data will be analyzed, and what will be done with the 
results as well as how the privacy and personal freedom of the 
public will be protected by the process itself. We expect full 
and complete disclosure.
    In 2 weeks after gathering and evaluating the information 
that will be presented today, the subcommittee will reconvene 
and examine this issue from a standpoint of privacy and 
personal freedom concerns in part 2 of this hearing. The 
subcommittee believes this is a good place to start from. From 
an oversight perspective, we look forward to working with these 
agencies as they continue to plan and implement their proposals 
for enhancing homeland security.
    [The prepared statement of Hon. Adam H. Putnam follows:]

    [GRAPHIC] [TIFF OMITTED] T0399.001
    
    [GRAPHIC] [TIFF OMITTED] T0399.002
    
    Mr. Putnam. It's my pleasure now to yield to the gentleman 
from Missouri, the ranking member, Mr. Clay, for any opening 
remarks that he may have.
    Mr. Clay. Thank you, Mr. Chairman. And thank you for 
calling this hearing. I look forward to today's testimony. When 
government agencies collect information about American 
citizens, then those citizens have a right to see that 
information and correct it if there are errors. I would like to 
begin by quoting one of our witnesses from our last hearing on 
data mining, Professor Jeffrey Rosen. At that hearing, 
Professor Rosen opened his testimony with this statement: 
``It's possible to design data mining technologies in ways that 
strike better or worse balances between liberty and security. 
But there is no guarantee that the executive branch or the 
technology just left to their own devices will demand and 
provide technologies that strike the balance in a reasonable 
way. Congress, therefore, has a special responsibility to 
provide technological and legal oversight of data mining to 
ensure that the most invasive searches are focused on the most 
serious crimes.''
    Our job today is to gather as much information as possible 
about these three programs so that we can assure that the 
balance between liberty and security is a good one. Over the 
past 2 years, we have seen a heavy thumb on the balance scale 
in favor of security. However, it is not clear that we are 
necessarily more secure because of it. We have also seen the 
liberty of individuals abused in ways we have not seen in this 
country since the internment of the Japanese during World War 
II. We learned in hindsight that breach of liberty was a 
terrible misuse of government power. The government quietly 
admitted so when it turned to those people interned and asked 
them to serve in the military or to work as translators.
    Much later our government officially apologized. President 
Clinton, in issuing that policy said, ``we recognize the wrongs 
of the past and offer our profound regret to those who endured 
such grave injustice. We understand that our Nation's actions 
were rooted in racial prejudice and wartime hysteria. And we 
must learn from the past and dedicate ourselves as a Nation to 
renewing and strengthening equality, justice and freedom.''
    Today our government faces a threat to our national 
security that many have compared to World War II. President 
Bush compared the attack on the World Trade Center to the 
bombing of Pearl Harbor. In the days that followed that attack, 
the President's speech writers used President Roosevelt's 
speeches from December 1941 to shape President Bush's remarks. 
We must learn from the past and not allow our fears to destroy 
the very liberties for which we fight. The descriptions of the 
programs we are considering today with secret filings and 
warrantless searches of our electronic lives puncture that thin 
wall between liberty and security. At the same time, these 
programs have not proved that they have a benefit strong enough 
to justify that breach.
    Finally, I'd like to thank the Defense Advance Research 
Project Agency for providing this testimony to the subcommittee 
in a timely fashion. It's a shame that the FBI and TSA did not 
show the same respect for this subcommittee. Again, let me 
thank the witnesses for their testimony. And I ask that my 
statement be included in the record. Thank you, Mr. Chairman.
    Mr. Putnam. You're very welcome. Thank you, Mr. Clay. We 
appreciate your interest.
    [The prepared statement of Hon. Wm. Lacy Clay follows:]

    [GRAPHIC] [TIFF OMITTED] T0399.003
    
    [GRAPHIC] [TIFF OMITTED] T0399.004
    
    [GRAPHIC] [TIFF OMITTED] T0399.005
    
    [GRAPHIC] [TIFF OMITTED] T0399.006
    
    Mr. Putnam. And obviously this is going to be an 
interesting hearing. At this time, I'd like to recognize the 
vice chair of the subcommittee, the gentlelady from Michigan, 
Mrs. Miller.
    Mrs. Miller. Thank you, Mr. Chairman. Just a brief opening 
statement if I may. I look forward to hearing the testimony of 
all the witnesses. I'm very appreciative of all of you coming 
here today. I think this will be a fascinating hearing. I think 
the issue of factual data analysis as a tool to strengthen 
national security is certainly one of the most significant 
issues facing our Nation. In fact, today in our society. And 
with the implementation of the E-government Act of 2002 and the 
growing importance of information technology in our world 
establishing investigative techniques such as factual data 
analysis are vital, absolutely vital if our Nation is to 
successfully prosecute the war on terror.
    As we know, the terrorists seem to have an uncanny ability 
to adapt to our methods of prevention. In many instances, they 
are using our freedoms against us. In my view, I think we need 
to focus all of our resources and attention to ensure we are 
always at least one step ahead of them if possible. Federal 
officials currently have at their disposal the resources and 
knowledge to implement systems that assist us in this process. 
These officials currently occupy the vanguard of our defenses 
and need not to necessarily be hampered by bureaucracy in their 
efforts.
    However, the American people must have confidence that the 
Federal Government is using this new source of information in a 
very ethical and proper and effective way. It's absolutely 
essential that a proper balance be made between the operation 
of the government as it prosecutes the war on terror and the 
disclosure of operations to citizens that it was set up to 
protect. And for this reason, it's the responsibility of this 
subcommittee to ensure that the factual data analysis as a tool 
be not abused.
    For example, I'm certainly very encouraged that the Total 
Information Awareness Project [TIA], is being conducted by the 
Defense Advanced Research Projects Agency [DARPA], and it has 
been relatively transparent. Why only in its beginning phases 
this program provides a hope that we can analyze the patterns 
of a terrorist to anticipate their next move? And some have 
expressed concern about programs such as these. But the mere 
fact that Mr. McCraw, Admiral Loy, and Dr. Tether have agreed 
to testify certainly shows that the Federal Government is 
concerned about the perception that Congress and the public has 
about data analysis as well.
    I'm very much looking forward to working with the chairman 
and members of the subcommittee and full committee to ensure 
that this program receives the proper congressional oversight 
and I certainly will be interested to hear the testimony 
provided today. Thank you, Mr. Chairman.
    Mr. Putnam. You're very welcome, Mrs. Miller. We thank you.
    With that, we will move to the witnesses. You're all 
experienced with congressional testimony. You understand the 
light system. We'll ask you to adhere to the timing out of 
respect for everyone's schedules. Today each witness will 
testify on his own panel. After each witness has given the 5-
minute statement, the subcommittee will ask questions 
particular to that witness's agency. After all three panels 
have testified and answered this initial round of questioning, 
the three witnesses will return to the witness table to answer 
future rounds of questions.
    As you are aware, we swear in our witnesses. So if Mr. 
McCraw would please rise for the swearing in.
    [Witness sworn.]
    Mr. Putnam. Note for the record the witness has responded 
in the affirmative. I will ask if there are associates from 
your agency who intend to provide supporting evidence or 
testimony for the subcommittee, that you rise and be sworn in 
also at the appropriate time. Our first witness today is Steven 
C. McCraw, a 20-year FBI veteran. He's assisted this year to 
the newly created Office of Intelligence. His office will be 
responsible for implementing FBI intelligence strategies, 
making sure that intelligence is properly collected, managed 
and shared within the FBI, with State and local law enforcement 
through the 66 Joint Terrorism Task Forces, and with the 
intelligence community, including the new Terrorist Threat 
Integration Center.
    Previous to his current appointment, he was special agent 
in charge of the FBI San Antonio field office and served as the 
director of the Foreign Terrorist Tracking Task Force before 
that. We're pleased to have you and you're recognized for your 
statement.

   STATEMENT OF STEVE McCRAW, ASSISTANT DIRECTOR, OFFICE OF 
 INTELLIGENCE, FEDERAL BUREAU OF INVESTIGATION, ACCOMPANIED BY 
      WILLIAM HOOTEN, DEPUTY EXECUTIVE ASSISTANT DIRECTOR

    Mr. McCraw. Thank you, Mr. Chairman. First, I owe yourself 
and the members of the subcommittee an apology because you 
didn't have my statement well in advance. I have been notified 
that it has been cleared. I ask your permission that we do 
submit it for the record.
    Mr. Putnam. How quickly can you get it to the subcommittee 
using all the miracles of technology?
    Mr. McCraw. Well, I have to turn around here, Mr. Chairman, 
with your permission. Making copies and driving it here right 
now at this time.
    Mr. Putnam. I'm quite certain we have a fax machine. If 
they want to get it to us that way, we'll be able to afford the 
audience and others the opportunity to review it as well in a 
timely manner. We look forward to that. Thank you. You're 
recognized.
    Mr. McCraw. Thank you. First, I'd like to take the 
opportunity to thank each and every one of you for your support 
in enabling the FBI to modernize its information technology 
systems. I think, in fact, in all the statements that you made, 
you know, previously including the letter asking the FBI's 
participation in this important hearing, I noted the importance 
and value added benefits of utilizing technology. Clearly, the 
FBI's focus is trying to utilize these advances to manage and 
to find links, relationships, and patterns of individuals 
within its own data systems. To that end, it's the information 
in terms that the FBI legally and lawfully collects in the 
course of its investigations that becomes a part of its system 
of records.
    As background in terms of discussion of data mining, I 
believe it's important to understand the term ``data mining'' 
as it's used commonly today. It's defined as technology that 
facilitates the ability to sort through masses of amounts of 
information through data base explorations, extract specific 
information in accordance with defined criteria, and then 
identify patterns of interest to users.
    Also, as I mentioned before, it's an outstanding tool to be 
able to go through those data sets and identify links, 
relationships, and associations between individuals of 
interest. And in effect, what it does is automate what analysts 
and agents have had to do for years. So what it does, it allows 
analysts and agents to work cheaper, faster, smarter. And that 
is how the FBI, in terms of its Trilogy, is going about it.
    Now, one thing that has been critically important to the 
FBI and we have a strong commitment to and that is the rule of 
law, the Constitution, the statutes that you in Congress have 
passed, the Attorney General guidelines, the Privacy Act, and 
all of the laws and statutes that clearly delineate and 
guidelines for the FBI in terms of how they can properly and 
lawfully collect information. Because in effect, that's the 
information that we would be utilizing this technology on, its 
own internal information.
    One of the advantages, the FBI, from lessons learned over 
the years, is that we have, and the reasons we have what the 
Office of General Counsel has, an Administrative Law Unit, we 
have an Investigative Law Unit, the reason we have in the field 
the Chief Division Counsels so that these rules and regulations 
that are closely adhered to and followed up on is to ensure 
that an agent doesn't go out on an fishing expedition in terms 
of looking at an individual, arbitrarily looks at a person or 
surveils a person, but there's predication, a reason for doing 
it. It's the same thing that we're talking about in terms of 
data mining. There's a reason, first, to collect the 
information. Once we collect it lawfully, then naturally we 
want to exploit the latest technology so that we can work 
better, in protecting Americans from terrorism, from crime, and 
from foreign intelligence activities.
    Ensuring the appropriate controls to protect the privacy of 
the Federal Government data, we must also look at in terms of 
public source data. And that is data that is derived and sold 
by public companies that people have access to, you and I, 
certainly the private industry utilizes this, and clearly the 
FBI does utilize public source data as a tool, and clearly as a 
tool for leads.
    Again, so we can work more economically, we can save time, 
and we can be more efficient in what we're doing in terms of 
investigations. Now, we've learned from lessons learned, and as 
you well know, is that public source data is not always 
accurate. In fact, many times there are errors. So we have to 
be mindful it's a tool that requires followup investigation.
    I'm sure, there have been instances were they've come 
across erroneous information. I'm applying for a mortgage at 
this time--and they've identified inaccurate information. In so 
doing, it makes you mindful of how much other information was 
linked to a particular credit card inappropriately. Well, you 
know, the systems aren't perfect. They're run by people. That's 
why it's only a tool for FBI. And that when we use public 
source data that we extract the relevant components of it 
before we bring it into the FBI system of records. We don't, 
and we won't, go out and purchase wholesale data sets that are 
publicly available and incorporate that with the names of 
myself, my family, you and hundreds of thousands of other 
Americans in our system of records for convenience sake.
    Clearly we have an obligation to be mindful of those 
things. I look forward to any questions that you might have 
later. Thank you very much for your time.
    Mr. Putnam. Thank you, Mr. McCraw. I'm informed that we are 
either electronically or by fax receiving your testimony so we 
appreciate that.
    Mr. McCraw. We don't have the greatest success rate in 
technology, Mr. Chairman. I apologize for that. We're trying to 
get better with it.
    Mr. Putnam. We're going to try to help you.
    Mr. McCraw. Thank you, sir.
    [The prepared statement of Mr. McCraw follows:]

    [GRAPHIC] [TIFF OMITTED] T0399.007
    
    [GRAPHIC] [TIFF OMITTED] T0399.008
    
    [GRAPHIC] [TIFF OMITTED] T0399.009
    
    [GRAPHIC] [TIFF OMITTED] T0399.010
    
    [GRAPHIC] [TIFF OMITTED] T0399.011
    
    [GRAPHIC] [TIFF OMITTED] T0399.012
    
    Mr. Putnam. I'd like to recognize the gentleman from 
Missouri to begin his round of questions. Mr. Clay, you're 
recognized.
    Mr. Clay. Thank you, Mr. Chairman.
    Mr. McCraw, a few weeks ago, the FBI issued a final rule 
that exempted information held by the FBI like the information 
in Trilogy from the Privacy Act requirements that the 
information held on individuals be accurate and timely. In 
other words, the FBI is going to make no effort to assure that 
the information they hold on an individual is correct. Now, 
some of these records are available to local law enforcement 
officials. When someone is stopped for a traffic violation, the 
officer runs that person's identity through a number of 
systems, one of the systems checked is the FBI's National Crime 
Information Center. Under this new rule, the FBI no longer is 
obligated to assure that the information held by the National 
Crime Information Center is correct. That means that people 
will be detained and arrested based on inaccurate information 
when all they have done is roll through a stop sign. Will you 
explain to this committee why the FBI believes it is no longer 
necessary to verify the accuracy of the information it holds on 
individuals?
    Mr. McCraw. Mr. Clay, one reason without discussing the 
statute for years and maintaining these high standards in terms 
of what data is entered into the NCIC is for that very specific 
reason that you gave, is that we have police officers out 
there, they stop for a traffic ticket, run the name and they 
have been advised all of a sudden this person is wanted. In 
accordance with officer safety and established guidelines, they 
have this person arrested. Clearly, there has to be, you know, 
strict guidelines when you use NCIC, and in fact, it has to not 
be just for the FBI, but all users of NCIC that have access and 
enter records into it. Those requirements are, and it's been 
held in place and there will be no shifting of those 
requirements in terms of agents being obligated in terms of the 
accuracy of the information that goes in there.
    That's critical in terms of the FBI in terms of how it 
operates NCIC. It's also important for us in our own system of 
record that we have accurate information. But sometimes what we 
may find in our own system of records is a report, a lead that 
someone is suspected by somebody of doing something and that 
we're obligated as FBI, in the FBI, to followup on. And we find 
during the course of it that's not accurate.
    In fact we found that for other reasons that the allegation 
was made that wasn't accurate. I'm not in a position to discuss 
some of the technology issues or the statute that you referred 
to, but obviously, we want everything in our data base to be 
correct. But I can assure you there is instances where we 
collect information that when we do further investigation, we 
find out, in fact, that statement was not correct to begin 
with.
    Mr. Clay. So then what happens? Do you go back and correct 
your records? I mean, do you delete that information that is 
incorrect from the records or do you put it in a different 
category?
    Mr. McCraw. No, sir. We just include in the record that we 
have gone out and done this and determined that, in fact, this 
allegation was not true.
    Mr. Clay. So that's the system the FBI has in place.
    Mr. McCraw. Because we're obligated to keep all information 
that we collect and also show the business process of what we 
actually did or did not do.
    Mr. Clay. What about some information that is so inaccurate 
that you list some motorist stopped as being on the FBI 10 most 
wanted list, and then you detain this person and you find out 
he or she is not the right person. Then how do you correct 
that?
    Mr. McCraw. We're still obligated to report the facts and 
the facts are that in your scenario that the FBI made a 
mistake. We still need to maintain the fact that we had a 
report, we acted inappropriately, we made a mistake and it's 
still there and it's documented.
    Mr. Clay. It would seem that out of efficiency to law 
enforcement you would go and clean up that error and take it 
out of that record so that the next law enforcement officer 
doesn't pull that information up.
    Mr. McCraw. I couldn't agree more, Congressman. As it 
relates to NCIC, and what State and local law enforcement have 
access to, absolutely. If there's a mistake that has been made, 
or if someone has been located they have to be immediately 
taken out of NCIC because that issue could have been resolved. 
We have to be mindful and the rules and guidelines require that 
it is immediately corrected if there's an inaccuracy found at 
NCIC.
    Mr. Clay. And that's what happens now.
    Mr. McCraw. Absolutely.
    Mr. Clay. Thank you, Mr. Chairman.
    Mr. Putnam. At this time, I recognize the Vice Chair, Mrs. 
Miller.
    Mrs. Miller. Thank you, Mr. Chairman. Mr. McCraw, pleasure 
to have you here. First of all, let me just say that my 
experience with your agents in the Detroit area has been 
remarkable. You have some really fabulous folks there that have 
done a very excellent job. And I felt like I was achieving 
nirvana with them because I was speaking to them every single 
day after September 11 in my previous capacity as the Michigan 
Secretary of State, where we do the motor vehicle kinds of 
things.
    As you might recall, after September 11 there were nine 
individuals that appeared on every newspaper in the Nation 
where these fellows had--this is about a week after September 
11--these individuals had obtained commercial driver's licenses 
with hazardous material endorsements. And they were all 
ostensibly from Michigan. As we found out later, I think there 
were only two that actually got a CDL with a HAZMAT endorsement 
from us. The rest weren't Americans, but they were here of 
Arabic descent that had just gotten a driver's license through 
the sources that they should have not from another State, 
actually.
    But at any rate, you know you look at some of these things. 
And that actually led us to make a proposal. I wasn't a Member 
of Congress then, but something that we had talked about 
actually did become a part of the Patriot Act in my State, if 
you want a concealed weapons permit, we do a criminal 
background check. But if you want a commercial driver's license 
with a hazardous material endorsement and drive around with 
10,000 gallons of liquid propane, no problem, just fill out a 
form.
    So we thought then to try to think like terrorists 
ourselves. As I said in my opening statements many times, these 
individuals are using our freedoms against us.
    And I sort of preface that with asking you, if you were 
aware if the Federal motor carrier division has promulgated 
rules or implemented them in which there is a requirement for 
an FBI criminal background check for anyone who is receiving 
not a CDL, Commercial driver's license, but a hazardous 
material endorsement. We have a large population of people who 
are of Arabic descent, and I compliment the FBI and Justice. It 
is very common knowledge there was a large group of individuals 
who were called in to be questioned in the Detroit area, and I 
think it was handled with a high degree of sensitivity by the 
FBI. I certainly again commend the Detroit agents for how they 
handled that.
    But do you have any knowledge if anything is happening in 
that area where there is a criminal background check now 
required for those that are getting those kinds of things? What 
has your experience been as you are creating these data bases 
as you interact with State agencies such as a DMV or others?
    [The prepared statement of Hon. Candice Miller follows:]

    [GRAPHIC] [TIFF OMITTED] T0399.013
    
    [GRAPHIC] [TIFF OMITTED] T0399.014
    
    Mr. McCraw. First, I want to thank you for your kind words. 
I appreciate that. I'll relay that to the special agent in 
charge in Detroit, and hopefully to his agents as well. I'm not 
familiar with what has been done right now in terms of 
regulations. And it may be better asked to the TSA, they may 
know better. If I'm not mistaken I think that's in their 
bailiwick, by statute. I do know one of the things that the FBI 
has been criticized for over the years, that we've been trying 
to correct, and clearly the director stepped ahead in doing so 
in your question in terms of how we're using data bases. We 
take--and also addresses one of Mr. Clay's concerns as well--
individuals under investigation for terrorism that we actually 
have a predicated subject, and they're investigated by the full 
field investigation, the FBI has taken, the Director ordered 
and has taken those names and put them into NCIC.
    For two reasons, there is predication, there is not an 
arrest scenario with it, but there is predication so the State 
and local officers have access to that information who are 
really the front lines of public safety and need that 
information.
    So those are the types of things that we're doing, and 
there's a number of other information sharing initiatives that 
really are technology-based, like a national alert system. We 
want to be able to reach out to those chiefs of police using 
the latest technology through their cell phones and PDAs and 
text messaging, and let them know, we're in the process of 
doing those similar type of things and even looking, the 
chairman noted earlier, at the importance, in post September 
11, of sharing information in ways that we can do it better and 
collocate investigative data, multi agencies, use encryption 
point-to-point over the Internet backbone to access information 
and provide details. I hope that was responsive to your 
question.
    Mrs. Miller. It was. Because I think it's so important, as 
you mention, as we saw since September 11, we've had 
individuals that have been picked up for routine traffic 
violations that were suspected terrorists. And the patrol in 
the black and white car didn't have the information or whatever 
as they're trying to share some of these data bases. My 
understanding is that the FBI currently has 31 different data 
bases, separate data bases which you're trying to combine under 
this Trilogy project. First of all, why do you have 31 
different data bases? How is it working as you try to notify 
some of those?
    Mr. McCraw. It's not working. Fortunately behind me, I have 
the Deputy Assistant Director, who we are fortunate to have in 
the FBI, whose job is to make those things work. And for years 
we've had antiquated systems, stovepipe systems, and certainly, 
Mr. Hooten is in a better state to describe the state of 
affairs he inherited, but clearly it was a detriment in terms 
of what we needed to do in our mission.
    So I have no defense for it. Clearly it was a problem. 
Certainly this Director recognizes the need for technology. And 
fortunately for us, and thankful to you that you've empowered 
people from the outside that come with the latest technological 
skills like Mr. Hooten, Mr. Lauer to come in and address this 
important issue.
    Mrs. Miller. Thank you. I don't have any further questions, 
but I certainly look forward to working with you and make sure 
you have the resources that you need to work these systems. 
It's critical. No use sitting here pointing fingers on what we 
should have done 3, 4 or 5 years ago. We need to look to the 
future. We've got a new enemy. These terrorists are different. 
They live in the shadows and prey on the innocent. We do need 
to utilize technology to assist you in doing your job. Thank 
you very much.
    Mr. Putnam. Thank you very much. We do have your testimony 
so thank you.
    Mr. McCraw. Again, apologize for the delay.
    Mr. Putnam. My understanding for the record, it was held up 
at OMB, not at FBI; is that correct?
    Mr. McCraw. I don't know that I'm supposed to comment on 
why it's held up. I know one thing, if I had done it sooner, it 
would have been likely cleared in time. So it's really my 
fault.
    Mr. Putnam. Well, we're glad that we have it now.
    In your testimony you indicated that you have traditionally 
used factual data analysis, the collection of data bases prior 
to Trilogy. Could you please compare what you have done in the 
past with the technology that Trilogy will provide for you.
    Mr. McCraw. Certainly. Currently in the FBI we have the 
system called ACS. It provides an antiquated software over a 
full text data that allows you to go through a number of green 
screens. It allows someone to try to do a full text query, just 
like you would a search on the Internet, a search engine. 
However, it is so cumbersome and is so difficult and you are 
overloaded with a tsunami of information that comes back hardly 
useful.
    Moreover, there is no visual link type of tools or link 
analysis tools that are common use; certainly the Department of 
Defense uses it; a number of different agencies have been using 
it successfully over the years.
    So right now the type of technology that is being brought 
on board and actually being used, even though development is 
being utilized right now by our counterterrorism analysts, is a 
tremendous benefit to our analysts. They will actually be able 
to go to specific sets of information for query. One of the 
major advances using a software package that actually works is 
that there is no question that we would have gotten to the 
Phoenix memo just asking the question, in terms of are there 
any threats in aviation.
    Also it's really important to utilize push technology, 
which the private industry has been using for years.
    Again, I'm just describing what we're doing now versus what 
I was able to do before. And let's say an analyst has a certain 
issue or topic, let's say it's ricin, anything in the FBI 
records gets loaded up into if it had ricin, whether it came in 
the community, whether it was an FD 302, which was an 
investigative report that the FBI did, or whether it was an 
insert or electronic communication, that information is pushed 
to the analyst.
    Now, the scope is the prototype, and Mr. Hooten and Mr. 
Lauer and others in their professional project managers are 
working on perfecting the technology. And it improves every 
day. Already we're seeing some tremendous advancements when we 
standardize the data, provide it in a useful format, and apply 
these state-of-the-art technology tools on top of it.
    Mr. Putnam. In quickly reviewing your written testimony, 
the term Trilogy is never mentioned. Would you define Trilogy 
for the subcommittee?
    Mr. McCraw. I think Mr. Hooten is probably better to define 
it, but Trilogy, I'll attempt it, my understanding is that it 
is the entire modernization of information technology in the 
FBI.
    If you don't mind, could we have Mr. Hooten.
    Mr. Putnam. Please stand.
    [Witness sworn.]
    Mr. Putnam. Note for the record the witness responded in 
the affirmative.
    Mr. Hooten. Trilogy is two very specific contracts. One is 
to redraw our infrastructure of networks, which is virtually 
non existent. That part has already been done. The second part 
is going through and upgrading our hardware inlcuding our old 
PCs. The third part is several applications of software, the 
main one is the virtual case file, which is the replacement, as 
Steve said, for the old ACS system. So it will be our new 
system of records. It's the management of our cases. But it's 
not a data analysis tool which is what this particular subject 
is. That system that Steve has been referring to is SCOPE, 
which is a development system that we're currently working on 
that's made up of the series of COTS products that we are just 
basically buying off the shelf and doing some quick 
modifications so that these analysts can have something in 
their hands that they can use right away. That is closer to 
this sort of data mining idea, going through sets of data, 
multiple data bases and looking for particular things.
    Mr. Putnam. The SCOPE would be more of what you would 
traditionally define as data mining than Trilogy?
    Mr. Hooten. Yes.
    Mr. Putnam. And SCOPE stands for what?
    Mr. Hooten. I was afraid you would ask me that. I can't 
tell you off the top of my head. We've been calling it SCOPE so 
long I forget what it stands for. I can find out for you 
though.
    Mr. Putnam. OK. That would be helpful. What new data bases 
would be searched through SCOPE or through the new Trilogy 
program that are not currently accessed or utilized today or 
prior to the deployment of those two programs?
    Mr. Hooten. Nothing new. It's the same thing we're doing 
now. The first one is our ACS, which is our system of records, 
that's our main data base. The nine other things that are very 
helpful to the analyst are called SAMnet, which are all the 
cables coming in.
    Mr. Putnam. According to your written testimony, ``the FBI 
uses information collected by public source companies to obtain 
information on individuals during the course of its terrorist 
criminal and foreign intelligence investigative activities.'' 
What type of public source companies have data bases that are 
accessed prior to an event that would trigger an investigation?
    Mr. McCraw. That we would use as an investigative tool? To 
name some of the public names, LEXIS/NEXIS, Choice Point. 
There's several of them out there that have information, 
driver's license information, government information that 
they've purchased and that through a query over the Internet 
and for a fee, you're able to find out additional information 
about a name. Again, it's a nice tool. It saves valuable lead 
time. But it has to be done not on a fishing expedition, it's 
done based upon a reason. There has to be a reason why you 
decided to run somebody through that data base.
    Mr. Putnam. Could you please elaborate some on what role 
these improvements, and you've outlined two or three different 
programs, how will they contribute to better collaborative 
efforts between the FBI and the CIA with the Terrorist Threat 
Investigation Center?
    Mr. McCraw. Well first and foremost, it allows the FBI to 
properly manage its information so we can extract the essential 
elements of information and to get that in through reports 
officers, and subject matter experts to get that to the TTIC, 
also to the Counterterrorism Center as well and to other 
customers out in the community that need that information.
    From the technology standpoint, Mr. Hooten can explain a 
lot better. It has been standardized so when there's sets of 
information that the FBI is legally able to provide the 
intelligence community with, it can do in a standardized 
format, that it can then use without additional, you know, 
contractors having to rewrite the format. Better----
    Mr. Putnam. Before you do, I think, for the record, if you 
would, please give your full name and your position.
    Mr. Hooten. William L. Hooten. My position is Deputy 
Executive Assistant Director over at administration.
    Mr. Putnam. Thank you. I appreciate that. Thanks for your 
help.
    Mr. Hooten. Sure.
    Mr. Putnam. Mr. Clay raised some interesting issues about 
the accuracy. What is the level of sophistication of technology 
today that an Arabic name, for example, would be case sensitive 
or would certain persons who have the same name and perhaps 
even the same middle initial and perhaps even the same middle 
name, what level of sophistication is there to prevent people 
from being caught up in a mistaken identity?
    Mr. McCraw. Well, obviously, transliteration has been a 
problem that all of us face in the government in terms of 
names. I mean, Waheed Alshiri, I know of at least 14 different 
types in juxtapositions of the name itself and in public source 
data alone in which it it appeared. And many times there is 
insufficient data that you can actually make a determination 
that it was, in fact, that person. Because there is no date of 
birth, biographical data or other relational type of data that 
you can be assured it's that person. That's why it's careful, 
especially if your operating in the public and proprietary data 
bases, that there is always followup along those lines, and 
that it's properly characterized, that information.
    Again, a tool within our own system now that we're bringing 
on greater and advanced tools, there is varying degrees of 
software that has greater success in terms of discerning those 
differences, in providing a greater ability of analysts to be 
able to try to get the transliterations, the juxtapositions or 
incorrect spellings during the course of an investigation that 
it was captured.
    So clearly, technology has improved. It empowers the 
analyst and agent to do things that we couldn't do in the past, 
but it still requires followup work on every piece. And 
certainly it's an analytical judgment, an investigative 
judgment when you brought this information together.
    Mr. Putnam. Mr. McCraw, we're going to have to move to the 
second panel recognizing, of course, that everyone will be on 
the panel together as soon as we have gone through these 
individually. Somebody had to be first and you drew the short 
straw. So thank you very much for leading us off on this 
hearing. At this time we'll excuse the first panel and seat the 
second.
    Mr. McCraw. Thank you.
    Mr. Putnam. Admiral, are you ready?
    Admiral Loy. I'm ready to be sworn, sir.
    Mr. Putnam. Let me introduce you first. We appreciate you 
being here. And look forward to your testimony. Admiral James 
Loy is the administrator of the Transportation Security 
Administration. Previous to his service in this position, 
Admiral Loy was Commandant of the U.S. Coast Guard and served 
as the Coast Guard chief of staff from 1996 to 1998. From 1994 
to 1996, he was Commander of the Coast Guard's Atlantic area. 
His other flag assignments were as chief of personnel and 
training and commander of the 8th Coast Guard district. A 
career sea going officer, Admiral Loy has served tours aboard 
six Coast Guard cutters, including command of a patrol boat in 
combat during the Vietnam War and command of major cutters in 
both the Atlantic and Pacific Oceans.
    Admiral Loy graduated from the U.S. Coast Guard Academy in 
1964 and holds masters degrees from Wesleyan University and the 
University of Rhode Island. Certainly a very distinguished 
career serving our Nation. We look forward to your service at 
the new Department of Homeland Security. Please rise and I'll 
swear you in.
    [Witness sworn.]
    Mr. Putnam. Note for the record, the Admiral responded in 
the affirmative. Is there anyone with you that needed to do 
that also?
    Admiral Loy. I don't think so.
    Mr. Putnam. You don't need anybody to answer questions?
    Admiral Loy. We'll see, sir.
    Mr. Putnam. Very well. You're recognized.

  STATEMENT OF ADMIRAL JAMES L. LOY, DIRECTOR, TRANSPORTATION 
                    SECURITY ADMINISTRATION

    Admiral Loy. Thank you, Mr. Chairman and good afternoon. 
Congressman Clay. Good afternoon, sir. Thanks for the 
opportunity to discuss CAPPS II as a project with your 
subcommittee. Mr. Clay, I clearly got your message on time, 
sir, and we'll make sure that we follow that closely in the 
future. If I may, sir, I'll offer my written testimony for the 
record and simply try to emphasize a couple of important points 
with my oral testimony about this project, and then answer your 
questions if I may, sir.
    First, it's important to recognize that the existing CAPPS 
system is seriously flawed and in need of replacement. We've 
studied this system at great length and replacement is the 
right word. It's too broken in both concept and execution to be 
upgraded or repaired. Discussing its shortcomings in a public 
hearing, I believe, is a bit inappropriate, sir, but I would 
like to offer the committee a follow-on closed briefing if 
there is any interest in that after our time together today.
    My point here is that CAPPS II would be a huge security 
improvement. I believe of all the elements that we've put in 
place in designing a system of systems for aviation security 
first and for the rest of transportation of now and into the 
future, CAPPS II has the most potential to improve both 
security and customer service. Our goal is to simply keep 
foreign terrorists off airplanes. And CAPPS II is a key piece 
of our interlocking system of systems. And it's also very 
important to note, I believe, Mr. Chairman, that we are working 
hard with other detection and screening project owners in the 
new Department of Homeland Security to ensure good stewardship 
of the taxpayers' investment in all of these projects. We don't 
need redundancies; we don't need overlaps; but we do need gaps 
closed, and we're working very hard to see those goals come to 
closure.
    Second, the goals of CAPPS II are simply twofold and very 
basic: They are to radically improve the identification 
authentication of travelers and with that, improve the 
identification and detection of known and unknown foreign 
terrorists before they board an airplane, including those that 
associate with terrorists.
    Third, the Aviation Transportation Security Act directed 
TSA to focus on CAPPS, the current system, and any of its 
successor systems for improvement and to ensure that any such 
system would evaluate all would-be passengers before they 
board. That's precisely what we're doing. Our review to date 
again clearly indicates the requirement for a replacement 
system.
    Fourth, CAPPS II will be a limited risk assessment tool 
based on dynamic intelligence information about the activities 
of known terrorists and their associates. It will be run by TSA 
inside a compartmented government firewall and sensitive to 
changing intelligence assessments or the alert conditions, for 
example, as set by DHS. Think of it, if you will, as a 
theostat, for the risk assessment scores returned by the CAPPS 
II tool can be compared to higher or lower limits set by 
intelligence inputs for the day.
    Fifth, unlike the classic definition of data mining as 
outlined in your invitation to testify, where one searches 
through reams of data to detect or identify hitherto unknown 
patterns, CAPPS II will be a traveler-activated search where 
traveler-offered data elements provided to secure a 
reservation, name, address, phone number, and date of birth, 
will initiate the identification authentication score, as well 
as the risk assessment score. The search will be to determine 
if well-known patterns are found to drive that risk assessment 
score higher than the threshold acceptable for the security 
environment of the moment.
    The products of the search, the result and risk score and 
identification score will translate to a simple direction to 
the checkpoint at the airport. Green, have a nice flight; 
yellow, provide secondary screening before boarding; or red, 
refer to law enforcement as one whose risk score deserves law 
enforcement scrutiny, and in any case, who will not board that 
flight.
    Sixth, we have recognized the importance of privacy 
concerns and have conducted considerable outreach to the 
privacy and civil liberties arenas. At multiple day offsites, 
we have collaborated with a good number of those with privacy 
concerns because we have them as well. We have met with groups 
of privacy officers from the business community, with groups of 
stakeholders in the privacy arena, and with individuals who 
have a deep conviction about the fourth amendment. We've also 
met with congressional representatives, with Senators and their 
staffs, and have listened intently so as to develop a privacy 
strategy for CAPPS II that will be a strength of our system, 
not a weakness.
    Most recently, we have engaged the review of Ms. Nuala 
O'Connor Kelly, the newly established privacy officer at the 
Department of Homeland Security, to validate our commitment to 
doing this right. And as we speak, she is with our delegation 
to the EU to sort out international privacy concerns. And I am 
simply of the mind that we can design, if we put our minds to 
it, a solid program where security and privacy are 
complementary goals. That's why we designed our Federal 
Register notice to solicit the widest spectrum of comment 
possible. When our listening is done, we will reissue that 
notice based on what we've learned.
    Seventh, CAPPS II will not build data bases on U.S. persons 
permitted to fly; will never see the commercial data being used 
to authenticate identification; will not search medical records 
or criminal records nor see credit ratings, overdue bills or 
any such data to assess risk; will not generate new 
intelligence; and will not keep even the risk scores after 
travel is complete.
    Eighth, CAPPS II will be a serious resource allocation 
tool. It will allow us to better schedule Federal air marshals, 
to better schedule screeners, even the new Federal flight deck 
officers--``guns in the cockpit,'' if you will--so as to 
optimize the resources we have to throw against the security 
problems we face.
    Last, Mr. Chairman, CAPPS II will be counterintuitive in 
the sense that we will not be looking for the proverbial needle 
in a haystack. Rather, we will be taking that haystack off the 
needle, identifying those thousands and thousands of perfectly 
innocent travelers, opting them in, if you will, thereby 
leaving only those we can't evaluate as OK to be the subject of 
added scrutiny.
    Today about 15 percent of the travelers are dubbed 
selectees, and undergo secondary screening. We believe we can 
bring that percentage way down, and thereby make not only a 
significant difference in security, but also a significant 
difference in customer service as well.
    Thank you for your attention, sir. I look forward to your 
questions.
    Mr. Putnam. Thank you, Admiral.
    [The prepared statement of Admiral Loy follows:]

    [GRAPHIC] [TIFF OMITTED] T0399.015
    
    [GRAPHIC] [TIFF OMITTED] T0399.016
    
    [GRAPHIC] [TIFF OMITTED] T0399.017
    
    [GRAPHIC] [TIFF OMITTED] T0399.018
    
    [GRAPHIC] [TIFF OMITTED] T0399.019
    
    Mr. Putnam. As with our previous witness, our ranking 
member will be recognized first for questions.
    Mr. Clay. Thank you, Mr. Chairman. Thank you Admiral for 
being here today. I want to salute you for your leadership in 
the area of airline security, over the TSA and being willing to 
serve this Nation in that capacity.
    Admiral, airline security has a troubled history of racial 
profiling, even before the attack on the World Trade Towers. 
During the 1991 Gulf war, individuals with Middle Eastern names 
were forced off their flights despite the fact they were 
American citizens. One gentleman, an American citizen, whose 
parents were from Bangladesh, was told he should carry his 
passport to prove his citizenship.
    Last year the ACLU testified before Congress of dozens of 
such incidents, individuals discriminated against in airports 
or on airplanes based on race and heritage. The same people who 
oversaw the private contractors who provided discriminatory 
security are now designing new systems. What is TSA doing to 
prevent racial profiling from continuing in our air 
transportation?
    Admiral Loy. Mr. Clay, the design work associated with 
CAPPS II as the replacement for the existent CAPPS program in 
place today has a very clear specification: there will be no 
racial or gender profiling. We frankly don't believe there's 
any value in going in that direction at the other end of a 
security risk assessment. What one has to do with the other is 
simply unknown to us. So the design work here is to keep such 
things totally out of the picture by specifying in the contract 
that we don't go there. There is no reason for it either, and 
it's certainly totally out of both my personal and our 
organizational ethic.
    Mr. Clay. You mentioned in your testimony the random checks 
that occur, I was leaving Orlando a couple weeks ago, from the 
Chair's area, I had a one-way ticket back here to Washington 
and I had four ``S"s on my ticket. Quickly routed into special 
security, take off the shoes and all of that.
    Mr. Putnam. It's not racial, man, I was there with you.
    Mr. Clay. You did it, too. I mean, so now under CAPPS II it 
will be less and less of that?
    Admiral Loy. Absolutely, sir.
    Mr. Clay. How do you get picked?
    Admiral Loy. Today CAPPS is a rule-based system. It has 
been that way with no changes to those rules for a rather 
lengthy period of time. In the immediate wake of the tragedies 
of September 11, 2001, those rules were actually reinforced. I 
would rather not go into a public listing of those rules, but I 
can tell you that they are recognizable; they are 
compromisable; they are broken, sir. That is exactly the reason 
why CAPPS, as a system in place today, needs to be replaced.
    We will not have rules associated with the manner in which 
CAPPS II will do its work. We will take advantage of added 
pieces of information that we will ask of travelers. A 
condition of a reservation is that you no longer just give your 
name; you give your name, address, phone number and date of 
birth, and allow the extraordinary technology of today to give 
us a risk score associated with authenticating that 
identification.
    Mr. Clay. Thank you. Admiral, 2 weeks ago the Wall Street 
Journal ran an article on the problems created by the no-fly 
list. That article began with the story of Larry Mussara. Mr. 
Mussara is a retired Coast Guard Commander, the father of 
three, a local hero in Alaska for his daring helicopter rescues 
of stranded fishermen and mountaineers.
    But, every time Mr. Mussara flies Alaska Airlines, which is 
about once a month, he gets stopped, often missing his plane. 
This kind of error occurs because Mr. Mussara has a name 
similar to one or more names on the no-fly list.
    In CAPPS II, the TSA is going to use a number of private 
data bases to make these same kind of comparisons. Will that 
increase the chance of a missmatch like the one Mr. Mussara 
faces?
    Admiral Loy. No, sir. It will actually radically decrease 
the chance of a mismatch. I want to make it clear here, Mr. 
Mussara will not be singled out in the days when CAPPS II is an 
active program, though he is focused on time and time again, 
unfortunately, under the program that is in place today.
    Mr. Clay. Just to wrap up with you. Can you share with us 
how many names are on the no-fly list? Is that available?
    Admiral Loy. Again, sir, I think that I would prefer to 
tell you that in confidence or in private. I will be happy to 
do that. I can offer that CAPPS II compared to CAPPS as it is 
in place today is quantum levels better in both identification 
scrutiny, and in risk assessment across a watch list created 
predominately by the Justice Department in the Terrorist 
Tracking Task Force over the course of this last year. So we 
are talking hundreds on one hand, tens of thousands on the 
other.
    Mr. Clay. Under CAPPS II the kind of mistakes that Mr. 
Mussara encounters will not occur?
    Admiral Loy. That is exactly right, sir.
    Mr. Clay. Well, thank you very much for that.
    Thank you, Mr. Chairman.
    Mr. Putnam. The gentlelady from Michigan, Mrs. Miller.
    Mrs. Miller of Michigan. Thank you, Mr. Chairman. And to my 
colleagues, I have also been under selection numerous times. It 
seems, to go through the whole check.
    I have to tell you that I was on a flight on Friday night 
with six Members. I noticed all of the Republicans were being 
selected, but not the Democrats. So I don't think it is racial, 
it is partisan. That is my observation.
    Admiral, I appreciate your testimony and certainly your 
service to our Nation as well. I would like to ask a question 
as you talked to categorizing it as a risk assessment score, 
perhaps we can call it a threat score, what have you.
    Will it only be the TSA that would have that kind of 
information of a threat score? Will you be utilizing or 
sharing, or sharing any of this information with commercial 
entities? If you do intend to share any of it with commercial 
entities, how can your organization ensure that the commercial 
entities are not sharing this information when they should not 
be?
    Admiral Loy. Ma'am, we will not be sharing the risk scores, 
either with respect to identification or with respect to final 
risk, with anyone outside that firewall I described as being 
the break point between inside government and outside 
government.
    We will be enormously concerned about four or five privacy 
parameters that I believe are the framing elements of what we 
wanted to build our privacy strategy around. But particularly 
to your point, we will not be getting into data itself; we will 
be designing arithmetic algorithms that will be able to search 
those data bases.
    Our first effort will be to take those traveler-initiated 
pieces of information--PNR data--aggregate them, and send them 
to commercial data bases for the manipulation, if you will, 
that offers us back an identification authentication.
    Goal one is to be able to look travelers in the eye and 
have great confidence that they are who they claim to be. That 
is job one, to get us from where we are today, with a name-
based system only, with its potential for challenge, and this 
goes directly to Mr. Clay's question as well, and toward a 
system where we have great confidence that the person who is 
asking for this reservation is the person he claims to be.
    Second, armed with that, we will run that authenticated 
name against the government data bases. That will reflect for 
us a final risk score determination. That will only be shared 
with another law enforcement organization if the purpose for 
which CAPPS II has been met, should they or should they not be 
allowed to board that plane. And there is a recognition in that 
risk score that law enforcement attention is actually in order.
    Mrs. Miller of Michigan. Admiral, you said, ``we want to 
look them in the eye and make sure that they are the person 
that they are claiming to be.'' So, let me ask if you have any 
comment on using technology, the retinal scans, looking them in 
the eye. That is the best technology that we really have 
available today.
    I will tell you, as a frequent traveler, I would be happy 
to have presecurity clearance and look me in the eye and make 
sure that you do the retinal scan and let me through the lines 
rather than standing there forever.
    Do you have any comment about whether we ever get to that 
point? I recognize the privacy advocates are talking about 
that. But, I mean, I think it makes a lot of sense.
    Admiral Loy. Well, we are actually working with the privacy 
advocates at the table, on design work, both with respect to 
CAPPS II and with respect to another project that we have 
underway, the transportation workers identification credential. 
The notion there is biometrical, such that the identification, 
so beating the so-called identity theft issue is very much 
within our grasp.
    CAPPS II will not have a biometric base associated with it, 
but I think we are only months away from having as a foundation 
block CAPPS II on one hand and the TWIC program on the other. 
We will build a registered traveler program that will be 
biometrically based, that will seek those players who are 
willing to step forward to get the background investigation, 
and get it accredited in the form of a biometrically-based 
card, so that we can facilitate a quicker passage through the 
airport system.
    Everyone will always go through the basic screening, and 
then of course if you trigger an alarm, based on having gone 
through the magnetometer, whatever screening is required there.
    If we can facilitate that in a frequent flyer line, or in 
some fashion that we can work out, and I am very optimistic 
that we can do that with the airlines, because they see it as a 
great value as well, we will end up doing exactly what you 
described. I would like to call it a registered traveler 
program, rather than frequent, which is, of course, associated 
with airlines exclusively, if you will, or trusted, the obverse 
of which I am not too keen on, labeling people as untrusted 
travelers.
    So we are right there. That is exactly the design work that 
we are following.
    Mrs. Miller of Michigan. Well, call it whatever you want 
to. I would like to sign up when you get it available, please.
    You mentioned that the purpose of CAPPS is really, a 
primary purpose is obviously to improve identification 
processes and those kinds of things. You heard me mention 
earlier I had been a person that was in charge of DMVs.
    And, of course, the driver's license really has become sort 
of the critical foundation of establishing anyones identity, 
whether they are utilizing driver's licenses, whether they are 
utilizing State identification cards, what have you.
    Let me ask you to comment on something that Secretary Ridge 
has, I know, made some comments on. There has been a lot of 
discussion amongst all of the States about the possibility, the 
potential of having a nationwide driver's license, because 
currently the type of primary documents that are required by 
the individual States to establish identity have such a huge 
fluctuation it is unbelievable. We take a lot of pride in 
Michigan, we think that we have some of the more stringent 
standards in the Nation. Minnesota also has very stringent 
standards.
    But I have never been able to figure out why it matters 
whether you get a driver's license in Minnesota or Michigan or 
Tennessee, or what have you, and that you have all of these 
different requirements. It must be an unbelievable challenge 
for yourself, the FBI and others, looking at these driver's 
licenses that are often times issued with erroneous 
documentation or very little kind of primary identification 
requirements.
    I think people are sometimes startled to know that it is 
the rule rather than the exception that almost every State in 
our Nation must issue a driver's license or a State 
identification card to people that we know are here illegally, 
illegal aliens are getting these driver's licenses. By most of 
our State laws we are required to give those out.
    So I can't imagine what kind of impact that is having on 
the CAPPS program, and some of these others as you try to 
identify trusted travelers.
    Admiral Loy. Yes, ma'am. I think you are absolutely right 
on point. We have given up using that base as a means by which 
we can gain confidence that a person who claims to be whoever 
they claim to be is really that person.
    In the other work of the Transportation Security 
Administration across all modes of transportation, that is 
where the transportation workers' identification credential is 
going, because, among other reasons, we can't have faith in the 
systems you were just describing.
    The notion of a biometrically based transportation workers' 
credential, both for identification purposes and for access 
control purposes, is where we believe we need to go in the 
transportation system at large. And we are working on two 
prototype projects in that regard, one in the Philadelphia 
area, and one in the Los Angeles area over the next several 
months.
    So I reinforce your concerns, and let you know that we 
can't go there with comfort and have to design a better 
mousetrap. I would also offer, based on your question to our 
first panelists, that just last Friday, ma'am, we did issue an 
interim final rule on hazardous materials endorsements on CDLs 
in conjunction with FMCSA, as well as RSPA, the Research and 
Special Programs Administration in DOT, with Justice alongside 
in terms of making sure we have met that requirement that you 
described in the Patriot Act.
    And it does require a BI, which is exactly, I think, the 
question you asked.
    Mrs. Miller of Michigan. Thank you, Admiral. Thank you, Mr. 
Chairman.
    Mr. Putnam. You are very welcome.
    Admiral, several things about the mechanics of CAPPS II 
program. First of all, the international terrorist 
organizations have shown a remarkable agility in selecting a 
variety of different targets. And in response to September 
11th, I think that we have disappropriately focused our efforts 
on protecting airline safety at the expense of rail, passenger 
cruise ships and other potential threats. Will this same 
technology be deployed for rail and passenger cruise lines?
    Admiral Loy. The potential is very much there, sir. My 
notion is that CAPPS II is a phased kind of project. Our first 
goal needs to be to construct; I have analogized it to our 
closets at home. And our first challenge is to build the rail. 
And then how we develop and use the multiple applications that 
might come from the risk assessment engines that will be 
designed as the rail. I liken those to multiple hangers 
sequentially being put on that rod over time.
    And with not only the knowledge of, but the consent of 
oversight-responsible organizations, not the least of which, of 
course, is the Congress.
    So, yes, my charge from Secretary Ridge is to build a 
national transportation system security plan, not an aviation 
security plan. And it does go to aviation, it does go to 
maritime, and it is about rail, transit systems, highways, and 
pipelines. All of those, plus maritime and aviation, compose 
our national transportation system.
    My goal is to make sure that Secretary Ridge is not found 
with a weak link among any or all of those aspects of our 
system. And, of course, that is just one of the puzzle pieces 
he has to fit into his much bigger challenge across the rest of 
our homeland.
    Mr. Putnam. Do you currently have congressional 
authorization to deploy that beyond air travel?
    Admiral Loy. We do not. That is exactly why I said it would 
be enormously important for us to come back and think it 
through carefully, not only with authorizing committees, but of 
course to seek the appropriations necessary to make it happen.
    Mr. Putnam. And how would the technology detect or review 
or assign a threat score to suspect domestic terrorists?
    Admiral Loy. The process would be absolutely similar, sir. 
Armed with those four pieces of information, the system would 
first of all build that identification score. That score then, 
as part of a review of those government data bases, would allow 
us to assign that final risk score. It frankly doesn't matter 
whether it is a foreign terrorist, although that is what we are 
looking for. Our challenge to allowing a U.S. person on that 
plane would simply be based on the fact that their score had 
elevated beyond the threshold of going from yellow to red, and 
we would then allow an investigative effort to take place by 
the right law enforcement organization.
    Mr. Putnam. So it would then also detect persons who are 
not necessarily a threat to that airline, but who are wanted 
for some other crime?
    Admiral Loy. As you heard me say earlier, sir, we are not 
searching NCIC as part of the data that we are looking at. This 
is a very focused tool, designed not without potential to do 
other things, if authorized and challenged by the Congress to 
do so, but at the moment, we are charged with finding, in the 
aviation sector, foreign terrorists or those associated with 
foreign terrorists and keep them off airplanes. That is our 
very limited goal at the moment.
    Mr. Putnam. Certainly the additional hangers that would be 
in your closet would inevitably lead to the technological 
ability in detecting anyone on rail, seacraft, aircraft, who 
would be then reviewed in data bases, that would include any 
number of warrants outstanding for any number of crimes.
    Admiral Loy. The potential there is very real, sir. And 
frankly at the other end of the day, even as heinous as it 
sounds, the ax murderer that gets on the airplane with a clean 
record in New Orleans and goes to Los Angeles and commits his 
or her crime, is not the person we are trying to keep off that 
airplane at the moment.
    Mr. Putnam. Today.
    Admiral Loy. Yes, sir.
    Mr. Putnam. But, clearly circumstances could change?
    Admiral Loy. As I indicated, there are several issues here.
    First of all, Mission Creep, if you will, is one of those 
absolute parameters that the privacy community is enormously 
concerned about, and I am enormously concerned about. We will 
build such concerns into the privacy strategy that we will have 
for CAPPS II.
    On the other hand, over the course of time, with an airing, 
clearly with the oversight associated with not only the 
Congress but our continued collaboration with our privacy 
colleagues, there are changes that can be made. There can be 
additional hangers hung on the closet rod.
    Mr. Putnam. Let me--you mentioned that the score--the 
criteria that determine the score change, depending on 
different circumstances.
    Admiral Loy. Yes, sir.
    Mr. Putnam. Who sets the criteria? Who makes those changes 
based on other intelligence or other circumstances?
    Admiral Loy. Well, I would offer, sir, that it is the 
identification score that comes back first across that firewall 
from its mix in the commercial data bases that will be 
searched. Armed with that score, we then assess the risk and 
produce the final risk score, and that score is going to be 
probably the same, regardless of what is also happening at the 
same time.
    That process will run its course, and we will end up with a 
risk score for that traveler. In the meantime, if we are at 
alert condition blue or yellow or orange or red, there may be 
enormously different attention being paid to one thing or 
another.
    If the intelligence drift over the course of that past 
month or week or day or hour is being focused on an airline or 
an airport or a flight or such things as that, then we would 
have the ability to adjust that rheostat in such a fashion, if 
there is a score of--pretend it is 100 max, if it was a score 
of 94, if 1 day would find you in a yellow capacity as opposed 
to red, based on the focus of intelligence that day, the 
security environment, if you will, that is associated with the 
world in which we are living in.
    We believe it is enormously important for this system to 
recognize adjustments in the flow of intelligence across our 
daily desks and be able to do something about it. At the 
moment, nothing like that exists in the CAPPS system that is on 
the books today.
    Mr. Putnam. And all of that is based simply on name, 
address, phone number, and date of birth?
    Admiral Loy. That is correct. That is the only data; that 
is only four pieces of data that, A, the traveler will offer, 
that, B, TSA will aggregate, that, C, will go to the commercial 
review process to produce the identification authentication 
score.
    And when it comes then back into across that firewall to be 
assessed against our government data bases for the final risk 
score, there are no other pieces of data. We will only see 
scores, not data beyond what the traveler offers us.
    Mr. Putnam. It will be sensitive enough that if Mr. Clay 
purchases a ticket from Washington and uses his Washington 
address and phone number, and returns back from Missiouri using 
his Missouri address and phone number, the discrepancy alone 
will not flag him red?
    Admiral Loy. That is exactly the case. Because, first of 
all, if it ever occurs, his opportunity for redress is, again, 
one of those parameters in the privacy strategy that we will 
have for CAPPS II that offers an appeal system to challenge 
decisions that have been made.
    Now, to the degree he goes red on that flight or even 
yellow, he has every right to call us. We are going to 
establish an ombudsman--we are not going to call it that, 
passenger advocate, I think, is the phrase that we are going to 
use--whose purpose in life is going to be to take calls from 
people who feel that they have been misread by the system, and 
adjust accordingly.
    That person will be able to search the data and come to the 
right answer.
    Mr. Putnam. In a timely manner for him to make his flight?
    Admiral Loy. I can't say that it will be in a timely manner 
for him to make his flight. If it requires research that the 
person, that the passenger advocate has to do for us, that is 
not a promise that I can make today.
    Mr. Putnam. I have overstepped the bounds of time that I 
set for everyone else. I apologize. We will excuse this panel 
temporarily and swear in the third panel and then bring all 
three of you back forward.
    So at this time we will excuse you, Admiral. Thank you for 
your testimony. And we will welcome panel three.
    Admiral Loy. Thank you, Mr. Chairman.
    Mr. Putnam. Thank you, sir.
    [Witness sworn.]
    Mr. Putnam. Note for the record Dr. Tether responded in the 
affirmative.
    Dr. Anthony J. Tether was appointed as Director of the 
Defense Advanced Research Projects Agency on June 18, 2001. 
DARPA is the principal agency within the Department of Defense 
for research, development and demonstration of concepts, 
devices and systems that provide highly advanced military 
capabilities.
    As Director, Dr. Tether is responsible for management of 
the agency's projects for high-payoff innovative research and 
development. Prior to his appointment as Director, DARPA, Dr. 
Tether held the position of chief executive officer and 
president of the Sequoia Group, which he founded in 1996.
    He has served as chief executive officer for Dynamics 
Technology Inc., vice president of Science Applications 
International Corp.'s Advanced Technology Sector, and then vice 
president and general manager for Range Systems at SAIC.
    He spent 6 years as vice president for technology and 
advanced development at Ford Aerospace and has also held 
positions in the Department of Defense serving as Director of 
DARPA's Strategic Technology Office from 1982 through 1986, and 
as Director of the National Intelligence Office in the Office 
of the Secretary of Defense from 1978 to 1982.
    Prior to entering government service, he served as 
executive vice president of Systems Control from 1969 to 1978, 
where he applied estimation and control theory to military and 
commercial problems with particular concentration on 
development and specifications of algorithms to perform 
realtime resource allocation and control.
    Your mother must be very proud.
    Mr. Tether. Well, I haven't been able to hold a job for 
very long.
    Mr. Putnam. You do move around a lot. But you are not 
exactly slumming. We welcome you to the subcommittee, and look 
forward to your testimony, if we can understand it.

 STATEMENT OF TONY TETHER, DIRECTOR, DEFENSE ADVANCE RESEARCH 
             PROJECT AGENCY, DEPARTMENT OF DEFENSE

    Mr. Tether. Well, thank you very much. I would like to 
offer my written testimony for the record if I can.
    I am not going to really go through much of the written 
testimony, since you have had it, and have had a chance to 
review it. So I just want to make a few points.
    First of all, my testimony today is less on TIA, the Total 
Information Awareness Program, as a program and more addresses 
just a very small part of that program, the data mining part of 
it.
    The TIA itself is a much larger program dealing with 
collaborative technology, language translation, biometrics 
identification and so forth and so on. Now, on the other hand, 
you all will be getting a major report on May 20th, assuming 
that I can get it through all of the coordination that still 
has to go on, which will describe in great detail for you the 
Total Information Awareness Program, as a program. So you will 
have that available to you very shortly.
    Data mining, as sometimes used, more commonly used, refers 
to the clever statistical techniques which basically seek to 
comb through large amounts of data looking for previously 
unknown but useful possible patterns. And, as you know, it has 
been used commercially by pharmaceutical companies and so forth 
and so on.
    The problem is that this approach, while useful for coming 
up with correlations, does lead to many false positives and so 
forth and so on. Also, it typically requires that all of the 
data be centralized in one place for those algorithms to work. 
We are really not pursuing that technique. I want to really 
make that clear, primarily, because if you have done nothing 
but read the papers about TIA, you are thinking that we doing 
nothing at DARPA but just piling through tons and tons of data 
about people in the United States looking for possible 
wrongdoings, and nothing could really be further from the 
truth.
    Our approach basically starts with a hypothesis about 
attack scenarios. Given an attack scenario, we create a model, 
a model which basically says, if this is the attack scenario 
being carried out, these are the observables, these are the 
questions that if we asked them, that came up positive, would 
indicate to us that this attack scenario was underway.
    So basically what we end up doing, we spend a lot of effort 
and time, basically in creating a model, which ends up with a 
pattern, a pattern that indicates that model or attack scenario 
is true.
    We then take that pattern to the data base, and look in the 
data base to see if that pattern exists. Now, this allows us 
basically to really cut down, to narrow to scope on the data 
bases on the answers. Also, it also allows us to let the data 
bases remain distributed. We don't have to bring the data to a 
central location with this approach.
    Basically, this is not a new approach, one of the questions 
was, well, how do you know that your approaches are going to 
provide any security? This is really an approach that has been 
under development by DARPA and other agencies for many years. 
One of the examples is in image processing.
    When we receive an image, we can either have an analyst go 
through and look at every little pixel to see if a target is 
there, or we can develop a model that says, this is what a 
target looks like in this picture, and then have the algorithms 
go through the image to find if that target is there.
    And this is called automatic target recognition, very 
successful. We have used it for years, and quite frankly, have 
just recently used it in the last Iraq war very successfully.
    Privacy, however, is really a major concern to us. And from 
the outset of the TIA program, we have really worried about 
privacy. Now, we worried about privacy for perhaps a different 
reason. But it does transcend into the public. As you can tell 
from reading my resume, I have been around for a long time. And 
every time there has been an intelligence failure, it has never 
been because we didn't have the data. We have always had the 
data.
    And, I think, as you see from the hearings on September 11, 
it turns out we always had the data. The problem was that the 
data was distributed. The problem is that some of data was held 
by CIA. Some of the data was held by NSA. The problem was is 
that there was no method to have everybody collaborate to work 
on a problem to bring that data to be able to answer questions.
    One of the reasons that there is difficulty in doing that, 
is that the agencies try to protect their sources and methods, 
for good reasons.
    I mean, this is for really good reasons. SIGINT, signals 
intelligence, is a special case. It has special rules and 
regulations because signals intelligence is gathering 
information overseas, but it can accidentally pull in what is 
known as data about U.S. persons, which is not just people, but 
also corporations. So there is a regulation that prevents 
people from really automatically sharing raw data.
    So we really are worried about the privacy concerns from 
how do we develop a system which would allow this collaboration 
to take place with distributed data bases for everyone to get 
together virtually, work on a problem, yet be comfortable that 
the privacy of their data bases was going to be held.
    And we have spent a lot of time and a lot of money on that. 
One of the major things we also believe is that an audit 
technique has to be developed. In other words, in order for 
people to feel comfortable about somebody seeing a piece of 
data, they want to know, in my world, the DOD world, there is a 
technique known as ORCON, originator controlled data.
    This is an attempt by the originator to make sure the data 
is not not used for other than the purpose for which it was 
granted to the individual. We believe that you need to have an 
audit technology which attaches itself to the data so that 
everyone, from that moment on, will know who is looking at that 
data and where has that data gone. We are spending a great deal 
of money and effort trying to do that too.
    I think I will stop, because time is getting late and I ask 
you for your questions.
    Mr. Putnam. Thank you, Dr. Tether.
    [The prepared statement of Mr. Tether follows:]

    [GRAPHIC] [TIFF OMITTED] T0399.020
    
    [GRAPHIC] [TIFF OMITTED] T0399.021
    
    [GRAPHIC] [TIFF OMITTED] T0399.022
    
    [GRAPHIC] [TIFF OMITTED] T0399.023
    
    [GRAPHIC] [TIFF OMITTED] T0399.024
    
    [GRAPHIC] [TIFF OMITTED] T0399.025
    
    [GRAPHIC] [TIFF OMITTED] T0399.026
    
    [GRAPHIC] [TIFF OMITTED] T0399.027
    
    [GRAPHIC] [TIFF OMITTED] T0399.028
    
    [GRAPHIC] [TIFF OMITTED] T0399.029
    
    Mr. Putnam. We will lead off with Mr. Clay, again.
    Mr. Clay. Thank you, Mr. Chairman. And thank you, Dr. 
Tether for being here.
    You indicate that TIA will use transaction information held 
by private companies. What will you do to assess the accuracy 
of those information systems?
    Mr. Tether. Well, actually, I don't think we ever really 
said that TIA will use transaction information held by private 
companies. I know it has been said in the press. Our emphasis 
really has been on the data that is currently legally collected 
by our intelligence community, and also the counterintelligence 
community.
    Now, what we have done, I, personally, believe that all of 
the data that we really need to have in order to detect these 
attacks already exists in that legally gathered data, because 
of my experience.
    On the other hand, we have looked at, ``Is there other data 
that perhaps would accelerate the process of coming to a 
conclusion about whether an attack is going on?'' So we have 
had research ongoing to see if there was other data besides 
data that is normally collected by the intelligence community 
that could be used.
    Now, we have a lot of researchers. DARPA itself does not 
have any internal capability. We contract out. We have 
researchers and industry that really do the work. So we have to 
describe a problem to them to work on. And in some of the 
descriptions of the problems we talked about transaction data. 
And I am afraid that they probably took it maybe a little more 
literally than we really meant it. We didn't mean credit card 
data. We really were looking for people to research data as to 
what might be extra data.
    Now, we have found in that research that transportation 
data is probably data that makes a lot of sense to be included 
into the mix. Why? Because these terrorists have to travel. And 
they don't have their own means of communication, which means 
they are going to have to take commercial means of 
communication like airlines, trains, rental cars, rental trucks 
and so forth and so on. So it looks like, if we were to add 
data that looks like it must be very advantageous, we would say 
transportation data really looks like it would be advantageous.
    But, again as to how do we know the data is good, and I 
don't mean this as a cop-out, but at DARPA we develop the 
tools. You have heard about these other two organizations who 
want to have this capability, but they can't have the 
capability if someone doesn't develop the tools.
    We are the ones that basically develop the tools for them 
to use. And presumably we don't collect any data, but we are 
developing tools for them to use on the data. And the accuracy 
of the data is really on their nickel.
    Mr. Clay. Doctor, along these lines, you indicate in your 
testimony that you want to operate a system that is consistent 
with the spirit not the letter of the Privacy Act.
    Mr. Tether. That is correct.
    Mr. Clay. I don't quite understand how you are going to do 
that. Fundamental to the Privacy Act is the right of 
individuals to see the information held about them, and to 
correct that information.
    Since DARPA will be using data held in the private sector, 
it cannot give individuals the right of access and correction, 
to say that you are going to comply with the Privacy Act seems 
misleading.
    How will you address this fundamental conflict?
    Mr. Tether. Well, the fundamental conflict again is going 
back to what we do. We don't collect any data. We are not the 
people that collect data. We are the people that supply the 
analytical tools to those who collect the data. Now, we do 
worry about privacy. And, therefore, we are providing 
technology so that the people in the different agencies, as 
they share data with each other, can be comfortable that their 
sources and methods are not being compromised.
    We are developing new technology for doing that. For 
example, one of the major efforts we have in this privacy 
technology, we have Oracle as a contractor. Why Oracle? Well, 
as you know they are a great data base company. And, in fact, 
it is the only way we will get the transitioning of that 
technology is to have a company like that develop the 
technology and be comfortable with its use, to pull it in.
    But we really are not the ones that collect the data. And I 
really want to make that very clear. We are the ones that 
supply the tools to the people that collect the data, who by 
the way are operating in full cognizance of all of the 
regulations.
    Mr. Clay. Thank you for that.
    You, know, you talked in your testimony about us knowing 
the hijackers, knowing they were here 2 days after September 
11, having known Attorney General Ashcroft for the last 20 
years as my State AG, Governor and U.S. Senator, he is a guy 
that can't keep a job either, but he told me that we knew that 
the hijackers were in this country, we just had no real good 
system of tracking them.
    Now, we'll be able to track them under this new technology?
    Mr. Tether. We are developing a technology, which will 
allow the agencies involved to easily collaborate with each 
other and to be able to ask questions and create a model.
    For example, if one of the models was, and we knew that 
using an aircraft as a weapon was something on their minds, 
because of the Philippines in 1995. So that was a known 
technique. Now, imagine if we had a model created about how 
would somebody go and use an airplane as a weapon? And we had 
created all of the indicators that had to be true.
    Well, one would say, well, it is going to be hard to have 
somebody have a pilot who is being paid to fly that airplane 
actually fly it into a building. So that means that someone is 
going to probably have to take over the plane and learn how to 
fly it.
    So one question would be, do we have any unusual people 
learning aviation? OK. Imagine if we had a collaborative system 
like we are talking about, and there was a set of questions up 
there, and one of them is, has anybody come across anybody 
wanting to learn how to fly aircraft under unusual 
circumstances, like maybe they don't care if they land?
    If that system had been in effect, that question in that 
memo that the FBI talked about, would have popped up, and 
people would have looked at it and looked further into it. I 
really believe that. What we are doing is providing the 
technology to allow that to happen, not the collection of the 
data itself.
    Mr. Clay. Thank you, Doctor, for your answers. Thank you.
    Mr. Putnam. The gentlelady from Michigan.
    Mrs. Miller of Michigan. Thank you, Mr. Chairman.
    Dr. Tether, I think that you may have just answered my 
question here. I am listening to you talk about modeling and 
models, and so as opposed to doing sort of broad surveillance, 
as you are doing the construct for your models and using some 
of those things for investigative work, then you say you do a 
model and you find a pattern, then you match the pattern 
against the data bases.
    I was going to ask you for an example, a pattern of what. 
Could you give us a pattern of what? And perhaps you just 
answered that, by talking about someone going to a flight 
school and learn how to fly an airplane, didn't care whether or 
not he landed. That is a pattern of odd behavior.
    Mr. Tether. That is a pattern of odd behavior that would 
come out if you had a center which people from different 
agencies were able to collaborate and actually address various 
questions. Now, we used to do this in the cold war in Germany 
in Stuttgart, there was a center where there was like 250 
questions that were developed over the years for what would the 
state of the Soviet Union be if it were going to attack us?
    And there were a whole bunch of questions. Some of them 
were, where is the Soviet leadership? Are they still publicly 
visible? The CIA could go over with their HUMINT people and 
find out the answer to that question and come back and say they 
are visible, yes we know where they are, and never have to 
disclose their source. And that is what I meant about having 
the technique to protect the sources and methods.
    But, we have had that system in the past. And it has worked 
really well. It is just never been applied here. We firmly 
believe that, first of all, all of our agencies have these 
stovepipes. These are really based on culture. We don't want to 
get rid of the stovepipes. I don't think anybody wants to 
either, because there is a certain value to having a culture.
    But, what we are trying to do is develop the technology so 
those stovepipes can be punctured full of holes. While the 
culture remains, it allows people to cross communicate in a 
very easy way. That is really all TIA is about. The data mining 
part is really the easiest part. It is all of the going before, 
the collaboration technology and coming up with the models, 
which you then take the pattern to the data base.
    That part, while really not trivial, is really the easy 
part. It is really the before part that is what we are trying 
to do. I hope that helps.
    Mrs. Miller of Michigan. Yes. Thank you.
    Thank you, Mr. Chairman.
    Mr. Putnam. You mentioned that there was no transactional 
data contemplated, that was in the media, but that actually was 
not contemplated by DARPA?
    Mr. Tether. Except in a research way to assess what might 
be a sweet spot of new data that is not currently being 
collected by our intelligence and counterintelligence 
organizations.
    Mr. Putnam. You identified transportation data.
    Mr. Tether. I probably will regret that. But that seems to 
be a major part to the puzzle, because of the fact that these 
terrorists don't have their own Air Forces or Navy and have to 
rely upon commercial capabilities.
    Mr. Putnam. But you are stating here, on the record, that 
you don't contemplate monitoring credit card transactions, 
library card check-outs, video rentals.
    Mr. Tether. Contemplating is the word I am having trouble 
with, because we have a lot of researchers who are out there 
looking at what data might be useful.
    I personally would be extraordinarily surprised if video 
rentals, and such, would be some of that data. The only person 
I know of who has said that credit card data monitoring would 
be a good thing to do, actually was President Clinton at an 
address at UC Davis in the spring of 2002, because he talked 
about some facts about the hijackers with their credit cards 
being maxed out, and so forth and so on, which might have 
tipped you off that something is wrong about this person who 
has six credit cards that are all maxed out.
    But that is sort of a hindsight type of input, that I don't 
really know if it would be valuable in doing the predictive 
something.
    Mr. Putnam. I am not trying to back you in a corner. I am 
trying to help you guys fix a little bit of your PR problem.
    Mr. Tether. I am trying to contemplate.
    Mr. Putnam. We pay you to contemplate.
    Mr. Tether. You pay us to contemplate.
    Mr. Putnam. So there is no plans to deploy?
    Mr. Tether. There are no plans to deploy. Contemplate, is 
there a researcher someplace that we are paying at some 
university who is thinking.
    Mr. Putnam. Again, I am not trying to back you in a corner. 
I am trying to get out there what is not occurring.
    Mr. Tether. Thank you.
    Mr. Putnam. I will stop there in the interests of time, and 
let's go ahead and reseat everyone. Let's take a 2-minute 
recess and allow everyone to come forward.
    [Recess.]
    Mr. Putnam. Let's go ahead and reconvene. I want to thank 
Admiral Loy and Mr. McCraw for remaining with the committee. We 
deliberately gave everyone their own panel to focus on the 
unique aspects of each of those systems, which I think is an 
important distinction to draw, because they are unique and have 
different purposes.
    But, under the umbrella of general factual data analysis or 
data mining, I think it also makes some sense to bring everyone 
back and talk about collaboration and some of the other issues 
that are common to all three departments or agencies.
    So we will, in keeping with our tradition, allow Mr. Clay 
to lead off with questions for the entire panel.
    Mr. Clay. Thank you, Mr. Chairman.
    There is a dark cloud of secrecy that hangs over this 
administration. And the programs we are discussing today are a 
part of the reason that cloud is so dark. These agencies are 
creating surveillance systems, and they don't want to tell the 
American people how they work. Just this week U.S. News 
reported that there are a number of children among the 
detainees at Guantanamo. How many we don't know, but we do know 
that the Secretary of State has objected to the situation.
    TSA wants to develop a profiling system and doesn't want to 
allow the public to see or correct the information that is used 
to profile them. The FBI has decided that it no longer has to 
worry about the accuracy of the information it holds on people. 
DARPA wants to build yet another system to profile the American 
public.
    Security at any cost is not what either Congress or the 
American public wants. Systems that skirt the Privacy Act or 
try to wiggle under the fourth amendment are not in our best 
interest. As we saw, when the public learned about the TIA 
program at DARPA, they were outraged, and Congress put a hold 
on the funding for that program.
    I would like to ask each of you to explain to the 
subcommittee what you are doing to explain to the public what 
you are doing, and why you must keep it a secret.
    Admiral Loy. Mr. Clay, thank you very much for the 
opportunity. First of all, TSA has filed a Federal Register 
notice, probably with as wide a set of opportunities for 
comment as has ever been published. We have reached out 
deliberately to a large spectrum of the privacy community, 
those folks who would be the first to support the fourth 
amendment.
    We intend to continue to hold that ongoing conversation by 
holding public meetings about exactly what CAPPS II is all 
about. We are absolutely not putting together a profiling 
system that we will be ashamed of or fear offering to the 
public for comment. In fact, we are going desperately quickly 
in the other direction, to give every opportunity to the 
public, through public meetings, and through publicly stated 
records, with every intention of refiling a public notice when 
we have heard that conversation carefully and forged the 
privacy strategy associated with CAPPS II.
    So I take issue with the kind of question that seems shaped 
around a conclusion, when, in fact, the conversation has not 
even been continuously held yet.
    Mr. Clay. Well, Admiral, I shape it that way because in our 
last hearing, Mark Forman, the Federal CIO, indicated that TSA 
was not meeting its deadlines for providing information to OMB. 
That was the end of March. Has TSA provided OMB with all of the 
information that OMB has requested about CAPPS II since that 
hearing?
    Admiral Loy. Indeed we have, sir. What Mr. Forman is 
responsible for is making sure that the business case end of 
CAPPS II complies, not only with the procedures appropriate in 
any instance, but those that are sensitive in a project like 
you and I are speaking about.
    Indeed we have provided that information to OMB.
    Mr. Clay. Thank you. Dr. Tether.
    Mr. Tether. Well, first of all, DARPA is not developing a 
system to profile the American public. And I hope, maybe I have 
convinced you in our previous conversation and the testimony.
    Nothing could really be further from the truth. I mean, we 
aren't doing that at all. As far as secrecy, I can't think of 
anybody that has been more open about what we are doing. It is 
because of the nature of us. In order to develop this 
technology, we reach out for everybody that we can. 
Consequently, we advertise what we are doing. In this 
particular area, all of what we are doing is really 
unclassified.
    Now, the application of it and the use of it on the data, 
the data becomes classified. But the actual technology and all 
is unclassified. We had a conference last year. We had 2,000 
people there. We disclosed fully what we were doing. The press 
was there. We even here in the House last year, in the House 
Authorization Report, what we were doing was lauded by the 
report. As, ``Hey, this is really great stuff.''
    Now, so it hasn't been kept quiet. On the other hand, 
because of the deluge of comments that happened last fall, I 
knew what was going on at DARPA, I was reading the papers. I 
mean, my God, I finally called up my guys, and asked if they 
were doing what I thought they were doing? I really became 
worried myself.
    I can imagine that you all, with the inputs that you were 
getting from your constituents, I know about them because a lot 
of them have been forwarded to me to answer, that you must have 
been worried too.
    Some of them were so outrageous we were stunned, quite 
frankly. The one mistake that we at DARPA made is that we were 
so stunned by the outrageous comments that we didn't do 
anything about it for some time. We just watched it. And 
finally we woke up and got over here and tried to get the truth 
out.
    Now, on the other hand the Department of Defense, the 
Honorable Pete Aldridge, who I work for, said, ``Look, I know 
that what you are doing is OK, Tony. But I am going to do two 
things to make sure that everybody else believes that too.'' 
First of all, two boards have been created in the Department of 
Defense, an internal board and an external board. The internal 
board is really set up to really just review the DARPA TIA 
project. OK, to make sure that the DARPA TIA project is truly 
following all of the rules and regulations and so forth and so 
on.
    And we have certain outputs already from our internal 
Inspector General that says what we are doing is in compliance 
with all of the laws and regulations. On the other hand, they 
have also established an external board. Now, this external 
board is comprised of people from the outside. I didn't give 
you a list in my testimony, but I would be happy to do that, of 
people on the outside, some are privacy advocates, but all are 
well known, and understand the Constitution and the laws and 
regulations.
    And the purpose of that is not just for us. But you see in 
the Department, if you want to do experiments on animals, you 
go to an animal board. You say, look this is an experiment 
using animals. Can I do it? And the board says yes or no. If 
you want to use humans, same thing. What we don't have is a 
board that really does privacy things.
    If you go to the board and say look, here is a new 
information technique that I want to use. You know, can I do 
it? Is it violating any laws and regulations? This external 
board is supposed to be a board that people like us can go to 
and get someone to say yes or no on what we are trying to do. 
But we really have been open.
    We are not developing a system to profile the American 
public.
    Mr. Clay. Thank you.
    Mr. McCraw.
    Mr. McCraw. Yes, sir, Congressman.
    First, I would like to say, you know, a core value of the 
FBI is a rigorous obedience to the Constitution and the rule of 
law. And that there is no justification at all when the FBI is 
in search of or investigating or trying to enforce the law, 
violating the law. We must adhere to all proscribed laws by 
Congress, the attorney general guidelines.
    In maintaining, you know, the data in our systems, we need 
to apply to those laws and we will. And whatever laws and 
whatever rules that are applicable, the internal guidelines, 
the FBI will abide by those.
    Mr. Clay. Let me say in summary, Mr. Chairman, to the 
witnesses, that TSA won't tell us how CAPPS II works or what 
data goes into it.
    DARPA wants to keep the same kinds of secrets. The FBI 
issued a rule saying that it doesn't have to keep its records 
on Americans accurate and up to date. I would like each of you 
to go back to your respective agencies and figure out what you 
can do to help build public confidence in your activities 
through openness and then report back to this committee and 
tell us what you think would make this process a little easier 
on the American public.
    Mr. McCraw. If I may, I would like to go back and research 
the rule that you are referring to, because, clearly, I need 
to. I will do that. And I will get back to you.
    Mr. Clay. Get back to us. I would appreciate it. Thank you.
    Mr. Putnam. Thank you, Mr. Clay. And if you have additional 
questions at any time, just let me know.
    I really sympathize with the position that all three of you 
are in. Prior to September 11th, we would have town hall 
meetings and letters and phone calls where people were unhappy 
about red light cameras. And then people were unhappy when 
Tampa hosted the Super Bowl and they deployed a new face 
recognition technology that scanned the crowd and detected 
people who may be terrorists or who may be criminals. And 
people were up in arms. It was a very hot constituent issue.
    On September 12th, people were up in arms that we had not 
done more to surveil potential bad people, to track potentially 
bad people, to keep tabs on them or to prevent their entry into 
this country and so forth and so on.
    We are plowing new ground here. We are taking advantage of 
wonderful new technologies that offer the hope of greater 
national security, and the threat of greater intrusiveness into 
innocent lives. And we hold Mr. McCraw and Admiral Loy on a 
regular basis responsible for all a lot of those things, for 
the safety of Americans, and yet all of us as policymakers in 
the executive branch and the legislative branch and the 
judiciary, are all trying to find this new line to go along 
with this new technology.
    So we have this obligation to find this balance and to take 
the promise of the technology, without having inaccuracies, 
cause good people to be held, detained or held in suspicion or 
cause to disrupt their lives, whether it is as simple as 
missing a flight or as serious as being detained and questioned 
and perhaps even held for some charge that is false.
    So that is what we are attempting to do with this Phase 1 
of these hearings is exercise our responsibility to weigh in on 
where these lines are. All of these events, all of these 
terrorist incidents have highlighted new weaknesses and new 
gaps in our ability as a country to detect or prevent them.
    And one of them is really a lack of collaboration, it is 
not a technological challenge, it is a human capital challenge, 
and a couple of you have referred to that. You are a creature 
of the FBI culture, a very proud, rich tradition in law 
enforcement.
    You are a DOD animal, a creature of that community. And 
Admiral Loy, you and your guys are still figuring out your 
culture.
    Admiral Loy. Well, I came with one from 200 years of 
culture.
    Mr. Putnam. You certainly have a rich tradition from the 
Coast Guard, and, hopefully, that will influence the spotted 
zebra that we have created at Homeland Security. But, how can 
each of you address the challenge of overcoming the cultural 
obstacles to utilize this technology, to share the data, to 
break down the mistrust, and have this system, these systems 
when they are deployed in an efficient way, have them truly 
collaborative across the Federal Government and including State 
and local law enforcement, beginning with Admiral Loy.
    Admiral Loy. Sir, it is an enormous challenge. If you go 
back to even a founding father, the balance was being discussed 
even then. Franklin said, ``He who would give up even a moment 
of liberty for hours of safety deserves neither.''
    Our challenge is to build the culture you were just 
describing. It is to take a set of core values, I use the 
phrase that was just used a moment ago, and in our case at TSA, 
to bring together a composite of people from so many different 
walks of life, not only the Federal Government, but from 
throughout America as well, and compose a culture. That culture 
would be based on the Constitution that we all hold so dear, 
and we would be reminded on a daily basis that what was 
attacked on September 11, 2001 was not just the World Trade 
Center, it was not just the Pentagon, but it was the whole 
notion, idea and ethic of what America is all about. We should 
be in the business in a post-September 11 environment of 
designing systems that recognize that very Constitutional 
foundation that we are so enormously proud of as Americans.
    I also, at this particular point in my 1-year tutelage of 
this organization, have come to conclude that the Founding 
Fathers, also had mobility as one of the inalienable rights 
they were talking about. My challenge is to protect that 
particular notion for Americans who want to get up and go, 
wherever they want to go, with some security and comfort. This 
includes not only being able to get to their destination, but 
any system that will be imposed on them for either their 
benefit for security or for their benefit for safety will be 
such that they don't feel they have given up something in the 
way of privacy in order to be comfortable in the way of 
security.
    That is the challenge that we have on a daily basis; you're 
absolutely right. It's the execution of those things, day after 
day after day, and the aggregation of that which will represent 
the answer at the other end of the day as to whether we, in our 
generation, had our chance to get it right and did so. That 
calls for a commitment to the very basic precepts that have 
always guided legislators, judges and operators in the 
executive branch, to meet the needs of Americans.
    Mr. Tether. First of all, I think you're absolutely right, 
collaboration is the key. And that is really the technology 
that really needs to be developed that will allow people like 
the three of us to be able to remain physically where our 
particular offices are, but yet be able to go into a ``room'' 
and have no constraints on being able to talk to each other 
like we're at this table right now. We're working hard on 
developing that collaboration technology which we believe in 
and we're experimenting with it. We have experimental nodes at 
several places within the DOD, INSCOM, overseas, SOCOM, 
STRATCOM, JFCOM, these are all DOD locations. We are putting in 
these experimental nodes where the collaboration technology and 
other tools are being tried. I really agree with you.
    Now how do we make people comfortable with what we're 
doing? Well, hearings like this. This is really the right way 
to do it. I don't know how else to do it except talk about what 
we're doing. To talk about it where people will listen. I'm 
sure there's press here in the room. And you know, this is the 
way to do it. Hearings like this, congressional oversight.
    As you know, we had a major flap, over the last year, and a 
lot of good came out of that. Because we weren't doing anything 
wrong. But now I believe that with the report that's coming out 
and the conversations that we've had, a lot of other people are 
understanding what we're doing. I think they're starting to be 
more comfortable with it. At least I'm not getting as many 
letters as I was. But collaboration is the tool, the coin of 
the realm. And meetings like this and hearings like this are 
the way we're going to make people in the United States 
comfortable. That's all I have.
    Mr. Putnam. Mr. McCraw.
    Mr. McCraw. Yes, sir. First it gives me another chance an 
opportunity to thank you. For example, the Patriot Act, now 
enables us to share data with other agencies that we could not 
up until the passing of the Patriot Act, as well as certain 
modifications in terms of the Attorney General guidelines. I 
know it's been talked about a great deal about the cultures 
within agencies not sharing information, protecting 
information. And I'm not going to say that the FBI has had the 
best reputation in the past in terms of sharing information, 
but we've all worked with the American public and we've all 
taken an oath. And one core value we all share is we're all 
willing to sacrifice for our country. Like you, we all love our 
country.
    We cannot afford to have something happen like September 
11th, because we didn't share information as a result of 
parochial reasons. And, frankly, with technology and the type 
of things that the good doctor is working on, clearly it 
enables us, I think, collectively as a community, to leverage 
that technology so that we can find those links, relationships, 
and do what all of us want to do, and every American citizen 
wants to do and that's protect the United States from a future 
terrorist attack.
    Mr. Putnam. Let me change gears just a little bit with you, 
Mr. McCraw, because Admiral Loy and I had sort of an extended 
dialog over the long-term potential of the CAPPS technology. It 
is difficult for me to believe that at some point in the 
future, when hopefully the national tension has declined a bit 
over international terror and it has returned to good old 
fashioned crime, it's difficult to believe that we wouldn't 
take that technology and use it as a tool to capture someone 
who had kidnapped a minor, was attempting to bring them back 
across different States on an airline, essentially an Amber 
Alert plus, it's difficult to believe that you wouldn't take 
that same technology that's already paid for, that's already in 
place, that's already omnipresent and use it to capture felons, 
drug dealers, perhaps even a step further, people who owe child 
support, it's limitless when you already have that type of 
technology in place.
    So as a law enforcement officer and former field agent, 
wouldn't the temptation be you would want to go to the parents 
of someone whose child has been kidnapped and say we've sent 
out the alert; there's no station, no airport, no cruise 
terminal anywhere, no rental car agency, that they could go 
without being picked up?
    Mr. McCraw. Well, you know, clearly, we're excited about 
the technology because it doesn't just make us more effective 
in terrorism, you're right, it also makes us more effective 
with our own data in the internal data we collect when we're 
trying to investigate corporate fraud, gang-related violence, 
the sniper case, all of these particular things will enable us, 
whatever the agent is investigating, and certainly to the 
degree that there's things in place, technology that's out 
there that we can employ over other data sets, as long as it's 
legally permissible by law, by statute and attorney general 
guidelines then clearly we would love to leverage that and will 
leverage that not being unlike we use some of the technologies 
out there in the private sector today and information when we 
are using some of the public source data base that we talked 
about.
    Mr. Putnam. Admiral Loy, is there anything you want to add 
to that?
    Admiral Loy. Yes, sir. I will just add that is what the 
notion of oversight is all about, both the general 
congressional oversight of what the executive branch is up to, 
checks and balances in the classic sense, but more 
appropriately, an oversight group not only about the Congress 
but to include folks who have as a raison d'etre in life, the 
fourth amendment, to have them be part of that system. So 
Mission Creep is not part of what occurs in a program without 
the attention very carefully and very publicly being brought to 
the attention of whoever is responsible for that program.
    So if it's me responsible for CAPPS II 5 years from now 
when what you just described has occurred, I want my oversight 
panel to be composed in such a fashion that as soon as there is 
the appearance of Mission Creep in the CAPPS II system, someone 
will very clearly bring that to my attention, and, as 
appropriate, bring it to your attention.
    Mr. Putnam. Dr. Tether, is it your opinion that the 
appropriate deployment of CAPPS II and Trilogy will 
substantially improve national security?
    Mr. Tether. Boy, that's a good question. The word is 
substantially. I believe I do know about CAPPS II. We were 
involved with the source election in CAPPS II, so I know about 
it. And I also do know something about Trilogy, because we were 
involved in talking with the FBI also about providing them some 
capability. We're prohibited by the way from providing either 
of these folks these tools that we're developing.
    And I would say yes. I believe that. Especially if they 
have the right tools. I'm hesitating because I'm not really 
sure what tools they're going to have. I really think the fury 
over what we're doing hasn't been by the people. The privacy 
people are well-intentioned, but they know that we're not the 
problem. It's the people who are going to use the tools and the 
people who are going to authorize the use of the tools that are 
the problem. But if they can stop the tools from being built, 
they don't have to worry about the other people. I really 
believe that's the reason for the fury over what we're doing. 
But the tools are important. These folks could use the tools. 
And I believe that they will. If he does his job right, with 
the right tools, security will definitely be greatly enhanced.
    Mr. Putnam. Mr. McCraw and Admiral Loy, will both of your 
programs, and Dr. Tether, as part of your research, does your 
program envision a mechanism for the public to have redress of 
inaccurate or incomplete information that would make them whole 
essentially for whatever harm is done?
    Admiral Loy. Absolutely sir. The passenger advocate role 
that I mentioned in my testimony is being designed to offer 
that redress opportunity for anyone aggrieved by the system.
    Mr. Tether. Same here. What we're doing is developing this 
audit technology so that the data, the information can only be 
used for the purpose for which it was granted and not abused by 
being taken out of that environment and used elsewhere. That 
audit technology again is very crucial to all of this. If it 
works well, the public or anyone can be assured that the data 
will only be used for what it was authorized to be used for, 
and not anything else.
    Mr. Putnam. Is it limited in its effectiveness to TIA, or 
does it have potential for these other----
    Mr. Tether. Oh, it's very general. In fact, it needs to be 
general because the data bases are not homogenous, they're 
heterogenous data bases so it has to be general.
    Mr. Putnam. Anybody can pull up their credit reports and 
find mistakes and have an opportunity----
    Mr. Tether. In our case, only if a person is authorized to 
see the credit report can he pull up the credit report. But 
then that credit report cannot be used for other purposes, 
without the audit technology either preventing it or raising an 
alarm that it's being done. That's what I mean.
    Admiral Loy. We're using, sir, a software system called 
Radiant Mercury, which is an NSA-accredited--literally, the top 
accreditation, if you will, that you can get in this regard. 
It's essentially a real-time logging system that will 
facilitate tracking for everybody that needs to understand 
access to the system that has been encountered. It will be a 
literally a real-time logging of who went in, got what, for 
what purpose. That's one of the auditing systems that we're 
using in the CAPPS II program.
    Mr. Putnam. Radiant Mercury?
    Admiral Loy. Radiant Mercury, a proprietary system by 
Lockheed Martin.
    Mr. Putnam. Only in homeland security?
    Admiral Loy. No, sir. That's software that any program 
manager who needs it for an auditing program or auditing 
software piece can either purchase or lease as appropriate from 
its proprietary owner.
    Mr. McCraw. Auditing is a very important function with us 
as well, including some of the risk technology that the doctor 
talked about. In fact, using that, we will definitely improve 
within the FBI information, who has a right to have access. Are 
there patterns of activity being done by an agent that would 
signal a concern in terms of the access? Also, we're even 
looking at using the same type of technology to provide greater 
oversight in terms of operations of sources. Because that 
pattern technology is there and how we can do a better job of 
using technology, building into Trilogy, things that can serve 
as triggers, as alarms of activity that we need to take a 
closer look at.
    Mr. Putnam. Dr. Tether, are you aware of other nation's 
approaches to data mining and where they are in their 
sophistication, and are there other customers in the Federal 
Government who have this level of data mining technology or 
greater that may be employed?
    Mr. Tether. You mean the technology that we're trying to 
develop?
    Mr. Putnam. Well, you've got three different ones 
represented. I'm just curious how many others there may be in 
the rest of the community or in other agencies that we're not 
as familiar with.
    Mr. Tether. With respect to other nations, I don't know of 
any. That may be just ignorance on my part. But I don't know of 
any that have technology that is anything like what we're 
talking about here. With respect to other Federal agencies, NSA 
obviously has a concern of people who have access to data that 
may have audit capability there. They're worried about making 
sure that data on U.S. persons stays in special compartments. 
CIA undoubtedly has the same with respect to their human 
capabilities. Other than that I don't think so. I mean, other 
than that, I think you've, for better or worse, you've got it 
here at the table.
    Mr. Putnam. Admiral Loy, any comments on that?
    Admiral Loy. I don't have any to add, simply because I'm 
not sure of what others might be. There is one point that you 
asked us to comment on earlier, Mr. Chairman. I would be remiss 
if I didn't add that the whole purpose of DHS, as I understand 
the legislation, was to facilitate the notion of information 
sharing, and that is exactly what's going on under Secretary 
Ridge. He is wedded clearly to the notion of the value in 
pulling together all those disparate agencies under one roof.
    For example, part of the review process for allowing CAPPS 
II to go on and making budgetary judgments with respect to 
building out our facility out at Annapolis Junction was to make 
absolutely certain that it took into acocunt other programs in 
the Bureau, the old Customs service, now the BCP Bureau under 
BTS or BICE, the new Bureau that used to be fundamentally INS 
under BTS, and whether there are screening projects going on in 
any of those organizations. Secretary Ridge and Under Secretary 
Hutchinson are making absolutely certain that we are present 
together in the same room and we're not allowed out until the 
notion of redundancies and overlaps and such are eliminated at 
the design stage and dealt with in a fashion that was the 
intention, I think, of the Congress when DHS was put together.
    Mr. Putnam. And you're satisfied that information sharing 
is occurring between the intelligence community?
    Admiral Loy. I still think we have a way to go. There's no 
doubt about that. But it is so much better than it was. We all 
had this quest for connecting the dots after September 11, 
2001, and as I think we heard from Dr. Tether the potential to 
connect those dots is infinitely better today than it was then, 
sir.
    Mr. Putnam. With regard to CAPPS II and the TSA, are we 
ahead of everybody else in the world? Do the Israelis have a 
model or someone else out there?
    Admiral Loy. Actually I'm going over to Israel at the end 
of the month with several purposes in mind, among which are to 
check the passenger prescreening process that they use there to 
make sure that we are learning about whatever they may have 
that they haven't shared with us already.
    Mr. Putnam. Very good.
    Dr. Tether.
    Mr. Tether. This is ignorance.
    Admiral Loy. Is Customs part of TSA?
    Mr. Tether. It's part of homeland security. Customs has a 
very good system for looking at imports coming in to be able to 
determine whether a cargo ship or cargo is suspect to know 
which ships to go and further inspect. And they have a very 
good system in place.
    Admiral Loy. If I may, Mr. Chairman, I have been personally 
writing about and convinced for several years that the point of 
origin, the point of destination, with transparency in between 
is the real key to our understanding the cargo piece, perhaps 
as well as the passenger piece, in terms of those who would be 
aboard or even the crews. But properly manifested cargo and 
people on ships at sea coming toward the United States is an 
enormous challenge for us to deal with. This goes to your 
earlier question about the rest of the transportation system 
not being left behind. As Americans, we have a pretty good 
penchant to continue to fight last year's war.
    Mr. Putnam. It's important that the rest of the world join 
us with all the ports that are now international and the 
tremendous volume of cargo.
    Admiral Loy. Commissioner Bonner has really done an 
excellent job in his outreach to ports around the world, in his 
so-called CSI initiative, which basically has convinced about 
25 of the major ports of the world who send things to the 
United States to be in a reciprocity agreement to allowing 
Customs, U.S. BCP inspectors, portal inspectors to be literally 
on the pier when things are being loaded on the ship as opposed 
to when they're already here. It's too late if they're already 
here.
    Mr. Putnam. Admiral, do you have any final comments for the 
subcommittee?
    Admiral Loy. Only to thank you for the opportunity as 
offered in the letter of invitation to offer to the committee 
some insights as to the intentions of where we're going and how 
we're going to do it. I appreciate that opportunity, sir, and 
look forward to working with you in the future.
    Mr. Putnam. Dr. Tether, anything from DARPA?
    Mr. Tether. No. Thank you very much. I do believe that 
hearings like this are the way to get the public comfortable, 
and thank you for having it. Thank you for having me.
    Mr. Putnam. Director McCraw.
    Mr. McCraw. There's only one director. If I want to keep my 
job, I better make that clear. The only thing I know it was to 
the culture question that has been echoed is that we recognize 
the FBI's unique collections of data, and we have to be able to 
share that data to the widest extent possible by law and also 
leveraging information technology that the good doctor has been 
working on over at DARPA. So thank you very much for the time 
of being here.
    Mr. Putnam. Thank you. I thank all of our distinguished 
witnesses for their participation today. I'm grateful for their 
cooperation with Congress. We're particularly grateful at your 
efforts to get in your written testimony. Your cooperation is 
key as we continue to deal with these important issues facing 
the Nation and the Congress. I want to thank the subcommittee 
members for their participation, particularly our ranking 
member. I appreciate everything that you've done. And in the 
event that there may be additional questions we did not reach 
today, the record shall remain open for 2 weeks for submitted 
questions and answers. Thank you all. The subcommittee stands 
adjourned.
    [Whereupon, at 5:21 p.m., the subcommittee was adjourned.]

