KLAYMAN v. OBAMA et al

Filing 22

MOTION For Leave to File Supplemental Declaration re 13 MOTION for Preliminary Injunction by LARRY E. KLAYMAN, CHARLES STRANGE, MARY ANN STRANGE (Attachments: # 1 Exhibit 1 -- Declaration of Professor Edward Felten, # 2 Exhibit 2 -- Washington Post Article, # 3 Text of Proposed Order)(Klayman, Larry)

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 1 of 35

UNITED STATES DISTRICT COURT
SOUTHERN DISTRICT OF NEW YORK

AMERICAN CIVIL LIBERTIES UNION;
AMERICAN CIVIL LIBERTIES UNION
FOUNDATION; NEW YORK CIVIL
LIBERTIES UNION; and NEW YORK CIVIL
LIBERTIES UNION FOUNDATION,
Plaintiffs,
v.
JAMES R. CLAPPER, in his official capacity as
Director of National Intelligence; KEITH B.
ALEXANDER, in his official capacity as Director
of the National Security Agency and Chief of the
Central Security Service; CHARLES T. HAGEL,
in his official capacity as Secretary of Defense;
ERIC H. HOLDER, in his official capacity as
Attorney General of the United States; and
ROBERT S. MUELLER III, in his official
capacity as Director of the Federal Bureau of
Investigation,

DECLARATION OF
PROFESSOR
EDWARD W. FELTEN

Case No. 13-cv-03994 (WHP)
ECF CASE

Defendants.

DECLARATION OF PROFESSOR EDWARD W. FELTEN
I, Edward W. Felten, declare under penalty of perjury, pursuant to 28 U.S.C. § 1746, that
the following is true and correct:
1.

The plaintiffs in this lawsuit have challenged what they term the “mass call-tracking”

program of the National Security Agency, and they have asked me to explain the sensitive nature
of metadata, particularly when obtained in the aggregate. Below, I discuss how advances in
technology and the proliferation of metadata-producing devices, such as phones, have produced
rich metadata trails. Many details of our lives can be gleaned by examining those trails, which
often yield information more easily than do the actual content of our communications.

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 2 of 35

Superimposing our metadata trails onto the trails of everyone within our social group and those
of everyone within our contacts’ social groups, paints a picture that can be startlingly detailed.
2.

I emphasize that I do not in this declaration pass judgment on the use of metadata

analysis in the abstract. It can be an extraordinarily valuable tool. But because it can also be an
unexpectedly revealing one—especially when turned to the communications of virtually
everyone in the country—I write in the hope that courts will appreciate its power and control its
use appropriately.
Biography
3.

My name is Edward W. Felten. I am Professor of Computer Science and Public Affairs,

as well as Director of the Center for Information Technology Policy, at Princeton University.
4.

I received a Bachelor of Science degree in Physics from the California Institute of

Technology in 1985, a Master’s degree in Computer Science and Engineering from the
University of Washington in 1991, and a Ph.D. in the same field from the University of
Washington in 1993. I was appointed as an Assistant Professor of Computer Science at Princeton
University in 1993, and was promoted to Associate Professor in 1999 and to full Professor in
2003. In 2006, I received an additional faculty appointment to Princeton’s Woodrow Wilson
School of Public and International Affairs.
5.

I have served as a consultant or technology advisor in the field of computer science for

numerous companies, including Bell Communications Research, International Creative
Technologies, Finjan Software, Sun Microsystems, FullComm and Cigital. I have authored
numerous books, book chapters, journal articles, symposium articles, and other publications
relating to computer science. Among my peer-reviewed publications are papers on the inference

2

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 3 of 35

of personal behavior from large data sets 1 and everyday objects, 2 as well as work on the
extraction of supposedly protected information from personal devices. 3
6.

I have testified several times before the United States Congress on computer technology

issues.
7.

In 2011 and 2012, I served as the first Chief Technologist at the U.S. Federal Trade

Commission (“FTC”). In that capacity, I served as a senior policy advisor to the FTC Chairman,
participated in numerous civil law enforcement investigations, many of which involved privacy
issues, and acted as a liaison to the technology community and industry. My privacy-related
work at the FTC included participating in the creation of the FTC’s major privacy report issued
in March 2012, 4 as well as advising agency leadership and staff on rulemaking, law enforcement,
negotiation of consent orders, and preparation of testimony.
8.

Among my professional honors are memberships in the National Academy of

Engineering and the American Academy of Arts and Sciences. I am also a Fellow of the
Association of Computing Machinery. A copy of my curriculum vitae is attached as Exhibit 1 to
this declaration.

1

Joseph A. Calandrino, Ann Kilzer, Arvind Narayanan, Edward W. Felten & Vitaly
Shmatikov, “You Might Also Like:” Privacy Risks of Collaborative Filtering, Proceedings of
IEEE Symposium on Security and Privacy (May 2011), http://bit.ly/kUNh4c.
2

William Clarkson, Tim Weyrich, Adam Finkelstein, Nadia Heninger, J. Alex Halderman &
Edward W. Felten, Fingerprinting Blank Paper Using Commodity Scanners, Proceedings of
IEEE Symposium on Security and Privacy (May 2009), http://bit.ly/19AoMej.
3

J. Alex Halderman, Seth D. Schoen, Nadia Heninger, William Clarkson, William Paul,
Joseph A. Calandrino, Ariel J. Feldman, Jacob Appelbaum & Edward W. Felten, Lest We
Remember: Cold Boot Attacks on Encryption Keys, Proceedings of USENIX Security
Symposium (August 2008), http://bit.ly/13Ux38w.
4

Federal Trade Commission, Protecting Consumer Privacy in an Era of Rapid Change:
Recommendations for Businesses and Policymakers (March 2012), http://1.usa.gov/HbhCzA.

3

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 4 of 35

The Mass Call Tracking Program
9.

On June 5, 2013, The Guardian disclosed an order issued by the Foreign Intelligence

Surveillance Court (“FISC”) pursuant to Section 215 of the Patriot Act (the “Verizon Order”). 5
This order compelled a Verizon subsidiary, Verizon Business Network Services (“Verizon”), to
produce to the National Security Agency (“NSA”) on “an ongoing daily basis . . . all call detail
records or ‘telephony metadata’ created by Verizon for communications (i) between the United
States and abroad; or (ii) wholly within the United States, including local telephone calls.” 6 The
Director of National Intelligence subsequently acknowledged the authenticity of the Verizon
Order. 7
10.

Following the disclosure of the Verizon Order, government officials indicated that the

NSA’s acquisition of call detail records is not limited to customers or subscribers of Verizon. In
particular, the NSA’s collection of this data encompasses telephone calls carried by the country’s
three largest phone companies: Verizon, AT&T, and Sprint. 8 Because these companies provide
at least one end of the vast majority of telecommunications connectivity in the country, these

5

Secondary Order, In re Application of the FBI for an Order Requiring the Production of
Tangible Things from Verizon Bus. Network Servs., Inc. on Behalf of MCI Commc’n Servs., Inc.
d/b/a Verizon Bus. Servs., No. BR 13-80 at 2 (FISA Ct. Apr. 25, 2013), available at
http://bit.ly/11FY393.
6

Id. at 2 (emphasis added).

7

James R. Clapper, DNI Statement on Recent Unauthorized Disclosures of Classified
Information, Office of the Director of National Intelligence (June 6, 2013),
http://1.usa.gov/13jwuFc.
8

See Siobhan Gorman et al., U.S. Collects Vast Data Trove, Wall St. J., June 7, 2013,
http://on.wsj.com/11uD0ue (“The arrangement with Verizon, AT&T and Sprint, the country’s
three largest phone companies means, that every time the majority of Americans makes a call,
NSA gets a record of the location, the number called, the time of the call and the length of the
conversation, according to people familiar with the matter. . . . AT&T has 107.3 million wireless
customers and 31.2 million landline customers. Verizon has 98.9 million wireless customers and
22.2 million landline customers while Sprint has 55 million customers in total.”).

4

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 5 of 35

statements suggest that the NSA is maintaining a record of the metadata associated with nearly
every telephone call originating or terminating in the United States.
11.

Assuming that there are approximately 3 billion calls made every day in the United

States, and also assuming conservatively that each call record takes approximately 50 bytes to
store, the mass call tracking program generates approximately 140 gigabytes of data every day,
or about 50 terabytes of data each year.
12.

Assuming (again conservatively) that a page of text takes 2 kilobytes of storage, the

program generates the equivalent of about 70 million pages of information every day, and about
25 billion pages of information every year.
13.

Members of Congress have disclosed that this mass call tracking program has been in

place for at least seven years, since 2006. 9
14.

On July 19, 2013, the day that the Verizon Order was set to expire, the Director of

National Intelligence disclosed that the FISC had renewed the NSA’s authority to collect
telephony metadata in bulk. 10
15.

As noted above, the Verizon Order requires the production of “call detail records” or

“telephony metadata.” According to the order itself, that term encompasses, among other things,
the originating and terminating telephone number and the time and duration of any call. Call
detail records also typically include information about the location of the parties to the call. See
47 C.F.R. § 64.2003 (2012) (defining “call detail information” as “[a]ny information that
9

See Dan Roberts & Spencer Ackerman, Senator Feinstein: NSA Phone Call Data Collection
in Place ‘Since 2006,’ Guardian, June 6, 2013, http://bit.ly/13rfxdu; id. (Senator Saxby
Chambliss: “This has been going on for seven years.”); see also ST-09-0002 Working Draft –
Office of the Inspector General, National Security Agency & Central Security Service (Mar. 24,
2009), http://bit.ly/14HDGuL.
10

Press Release, Foreign Intelligence Surveillance Court Renews Authority to Collect
Telephony Metadata, Office of the Director of National Intelligence (July 19, 2013),
http://1.usa.gov/12ThYlT.

5

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 6 of 35

pertains to the transmission of specific telephone calls, including, for outbound calls, the number
called, and the time, location, or duration of any call and, for inbound calls, the number from
which the call was placed and the time, location, or duration of any call”).
16.

Although this latter definition of “call detail information” includes data identifying the

location where calls are made or received, I will not address mobile phone location information
in this declaration. While senior intelligence officials have insisted that they have the legal
authority under Section 215 to collect mobile phone location information, they have stated that
the NSA is not collecting phone location information “under this program.” 11
17.

The information sought from Verizon also includes “session identifying information”—

e.g., originating and terminating telephone number, International Mobile Subscriber Identity
(IMSI) number, International Mobile station Equipment Identity (IMEI) number, etc. These are
unique numbers that identify the user or device that is making or receiving a call. Although users
who want to evade surveillance can make it difficult to connect these numbers to their individual
identities, for the vast majority of ordinary users these numbers can be connected to the specific
identity of the user and/or device.
18.

The information sought from Verizon also includes the “trunk identifier” of telephone

calls. This provides information about how a call was routed through the phone network, which
naturally reveals information about the location of the parties. For example, even if the
government never obtains cell site location information about a call, 12 trunk identifier

11

See Siobhan Gorman & Julian E. Barnes, Officials: NSA Doesn’t Collect CellphoneLocation Records, Wall St. J., June 16, 2013, http://on.wsj.com/13MnSsp; Pema Levy, NSA
FISA Metadata Surveillance: Is The Government Using Cell Phones To Gather Location Data?,
Int’l Bus. Times, Aug. 2, 2013, http://bit.ly/18WKXOV.
12

Cell site location information (“CSLI”) reflects the cell tower and antenna sector a phone is
connected to when communicating with a wireless carrier’s network. Most carriers log and retain
CSLI for the start and end of each call made or received by a phone, and some carriers log CSLI
6

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 7 of 35

information revealing that a domestic call was carried by a cable from Hawaii to the mainland
United States will reveal that the caller was in the state of Hawaii at the time the call was placed.
19.

In the present case, government officials have stated that the NSA retains telephony

metadata gathered under the Verizon Order, and others similar to it, for five years. 13 Although
officials have insisted that the orders issued under the telephony metadata program do not
compel the production of customers’ names, it would be trivial for the government to correlate
many telephone numbers with subscriber names using publicly available sources. The
government also has available to it a number of legal tools to compel service providers to
produce their customer’s information, including their names. 14
Metadata Is Easy to Analyze
20.

Telephony metadata is easy to aggregate and analyze. Telephony metadata is, by its

nature, structured data. Telephone numbers are standardized, and are expressed in a predictable
format: In the United States, a three digit area code, followed by a three digit central office
exchange code, and then a four digit subscriber number. Likewise, the time and date information

for text messages and data connections as well. Wireless carriers can also obtain CSLI by
“pinging” a phone whenever it is turned on, even if it is not engaged in an active call. The
precision of CSLI varies according to several factors, and “[f]or a typical user, over time, some
of that data will inevitably reveal locational precision approaching that of GPS.” The Electronic
Communications Privacy Act (ECPA), Part 2: Geolocation Privacy and Surveillance: Hearing
Before the Subcomm. on Crime, Terrorism, Homeland Sec. & Investigations of the H. Comm. On
the Judiciary, 113th Cong. (2013) (statement of Matt Blaze, Associate Professor, University of
Pennsylvania), http://1.usa.gov/1awvgOa.
13

See Letter from Ronald Weich, Assistant Attorney General, to Hon. Dianne Feinstein &
Hon. Saxby Chambliss, Feb. 2, 2011, http://1.usa.gov/1cdFJ1G (enclosing Report on the
National Security Agency’s Bulk Collection Programs for USA PATRIOT Act Reauthorization);
Siobhan Gorman & Julian E. Barnes, Officials: NSA Doesn’t Collect Cellphone-Location
Records, Wall St. J., June 16, 2013, http://on.wsj.com/13MnSsp.
14

See 18 U.S.C. § 2709 (national security letter); 18 U.S.C. § 2703(c), (d) (court order for
records concerning electronic communication service).

7

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 8 of 35

associated with the beginning and end of each call will be stored in a predictable, standardized
format.
21.

By contrast, the contents of telephone calls are not structured. Some people speak

English, others Spanish, French, Mandarin, or Arabic. Some people speak using street slang or in
a pidgin dialect, which can be difficult for others to understand. Conversations also lack a
common structure: Some people get straight to the point, others engage in lengthy small talk.
Speakers have different accents, exhibit verbal stutters and disfluencies. Although automated
transcription of speech has advanced, it is still a difficult and error-prone process.
22.

In contrast, the structured nature of metadata makes it very easy to analyze massive

datasets using sophisticated data-mining and link-analysis programs. That analysis is greatly
facilitated by technological advances over the past 35 years in computing, electronic data
storage, and digital data mining. Those advances have radically increased our ability to collect,
store, and analyze personal communications, including metadata.
23.

Innovations in electronic storage today permit us to maintain, cheaply and efficiently,

vast amounts of data. The ability to preserve data on this scale is, by itself, an unprecedented
development—making possible the maintenance of a digital history that was not previously
within the easy reach of any individual, corporation, or government.
24.

This newfound data storage capacity has led to new ways of exploiting the digital record.

Sophisticated computing tools permit the analysis of large datasets to identify embedded patterns
and relationships, including personal details, habits, and behaviors. As a result, individual pieces
of data that previously carried less potential to expose private information may now, in the
aggregate, reveal sensitive details about our everyday lives—details that we had no intent or
expectation of sharing.

8

Case 1:13-cv-03994-WHP Document 27

25.

Filed 08/26/13 Page 9 of 35

IBM’s Analyst’s Notebook and Pen-Link are two such computing tools. Both are widely

used by law enforcement and intelligence agencies for this purpose. 15
26.

IBM’s Analyst Notebook product is a multi-purpose intelligence analysis tool that

includes specific telephony metadata analysis features, which are “routinely” used to analyze
large amounts of telephony metadata. 16 IBM even offers training courses entirely focused on
using Analyst’s Notebook to analyze telephone call records. 17
27.

Pen-Link is a tool that is purpose-built for processing and analyzing surveillance data. It

is capable of importing subscriber Call Detail Record (“CDR”) data from the proprietary formats

15

Public Safety & Law Enforcement Operations, International Business Machines (last visited
Aug. 22, 2013), http://ibm.co/1avGItq (“IBM® i2® solutions help law enforcers to turn huge
volumes of crime data into actionable insights by delivering tools for tactical lead generation,
intelligence analysis, crime analysis and predictive analysis.”); see also Defense and National
Security Operations, International Business Machines (last visited Aug. 22, 2013),
http://ibm.co/18nateN (“IBM i2 solutions for military and national security organizations have
been used across the world to process and analyze the vast quantities of information that they
collect, to generate actionable intelligence and to share insights that help identify, predict and
prevent hostile threats.”); see also Pen-Link, Unique Features of Pen-Link v8 at 16 (April 17,
2008), http://bit.ly/153ee9g (“Many U.S. Federal Law Enforcement and Intelligence agencies
have acquired agency-wide site license contracts for the use of Pen-Link in their operations
throughout the United States…Pen-Link systems are also becoming more frequently used by
U.S. intelligence efforts operating in several other countries.”).
16

Case Studies: Edith Cowan University, IBM i2 Solutions Help University Researchers
Catch a Group of Would-Be Hackers, International Business Machines (Mar. 27, 2013),
http://ibm.co/13J2o36 (“Analyzing this volume of data is nothing new to many law enforcement
users who routinely analyze tens of thousands of telephone records using IBM® i2® Analyst’s
Notebook®.”).
17

Course Description: Telephone Analysis Using i2 Analyst’s Notebook, International
Business Machines (last visited Aug. 22, 2013), http://ibm.co/1d5QlB8 (“This intermediate
hands-on 3-day workshop focuses on the techniques of utilizing i2 Analyst's Notebook to
conduct telephone toll analysis…Learn to import volumes of call detail records from various
phone carriers, analyze those records and identify clusters and patterns in the data. Using both
association and temporal charts, discover how to use different layouts and more advanced tools
to analyze telephonic data quickly and effectively.”).

9

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 10 of 35

used by the major telephone companies, 18 it can import and export call data to several federal
surveillance databases, 19 as well as interact with commercial providers of public records
databases such as ChoicePoint and LexisNexis. Pen-Link can perform automated “call pattern
analysis,” which “automatically identifies instances where particular sequences of calls occur,
when they occur, how often they occur, and between which numbers and names.” 20 As the
company notes in its own marketing materials, this feature “would help the analyst determine
how many times Joe paged Steve, then Steve called Barbara, then Steve called Joe back.” 21

Figure 1: Screenshot of IBM’s Analyst Notebook. 22

18

See Pen-Link, Unique Features of Pen-Link v8 at 4 (Apr. 17, 2008), http://bit.ly/153ee9g
(describing the capability to import 170 different data formats, used by phone companies to
provide call detail records).
19

Id. at 4.

20

Id. at 7.

21

Id.

22

Image taken from Data Analysis and Visualization for Effective Intelligence Analysis,
International Business Machines (last visited Aug. 22, 2013), http://ibm.co/16qT3hw.

10

Case 1:13-cv-03994-WHP Document 27

28.

Filed 08/26/13 Page 11 of 35

The contents of calls are far more difficult to analyze in an automated fashion due to their

unstructured nature. The government would first have to transcribe the calls and then determine
which parts of the conversation are interesting and relevant. Assuming that a call is transcribed
correctly, the government must still try to determine the meaning of the conversation: When a
surveillance target is recorded saying “the package will be delivered next week,” are they talking
about an order they placed from an online retailer, a shipment of drugs being sent through the
mail, or a terrorist attack? Parsing and interpreting such information, even when performed
manually, is exceptionally difficult. To do so in an automated way, transcribing and data-mining
the contents of hundreds of millions of telephone calls per day is an even more difficult task.
29.

It is not surprising, then, that intelligence and law enforcement agencies often turn first to

metadata. Examining metadata is generally more cost-effective than analyzing content. Of
course, the government will likely still have analysts listen to every call made by the highestvalue surveillance targets, but the resources available to the government do not permit it to do
this for all of the calls of 300 million Americans.
The Creation of Metadata Is Unavoidable
30.

As a general matter, it is practically impossible for individuals to avoid leaving a

metadata trail when engaging in real-time communications, such as telephone calls or Internet
voice chats.
31.

After decades of research (much of it supported by the U.S. government), there now exist

many tools that individuals and organizations can use to protect the confidentiality of their
communications content. Smartphone applications are available that let individuals make
encrypted telephone calls and send secure text messages. 23 Freely available software can be used
23

Somini Sengupta, Digital Tools to Curb Snooping, N.Y. Times, July 17, 2013,
http://nyti.ms/12JKz1s (describing RedPhone and Silent Circle).
11

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 12 of 35

to encrypt email messages and instant messages sent between computers, which can frustrate
government surveillance efforts traditionally performed by intercepting communications as they
are transmitted over the Internet.
32.

However, these secure communication technologies protect only the content of the

conversation and do not protect the metadata. Government agents that intercept an encrypted
email may not know what was said, but they will be able to learn the email address that sent the
message and the address that received it as well as the size of the message and when it was sent.
Likewise, Internet metadata can reveal the parties making an encrypted audio call and the time
and duration of the call, even if the voice contents of the call are beyond the reach of a wiretap.
33.

There also exist security technologies specifically designed to hide metadata trails, but

those technologies do not work quickly enough to allow real-time communication. The general
technique for hiding the origin and destination information for an internet communication
involves sending data through a series of intermediaries before it reaches the destination, thus
making it more difficult for an entity such as a government agency to learn both the source and
destination of the communication. (Such data is conventionally encrypted so that the
intermediaries cannot capture it; and a series of intermediaries is used so that no one
intermediary knows the identities of both endpoints.)
34.

The most popular and well-studied of these metadata hiding systems is The Tor Project,

which was originally created by the U.S. Naval Research Lab, and has since received significant
funding from the State Department. One significant and widely acknowledged limitation of Tor
is the noticeable delay introduced by using the tool. Web browsing conducted through Tor is
much slower than through a direct connection to the Internet, as all data must be sent through a
series of Tor relays, located in different parts of the world. These volunteer-run relays are

12

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 13 of 35

oversubscribed—that is, the demands on the few relays from hundreds of thousands of Tor users
are greater than the relays can supply, leading to slowdowns due to “traffic jams” at the relay.
35.

Browsing the web using Tor can be painfully slow, in some cases requiring several

seconds or longer to load a page. Real-time audio and video communications require a
connection with minimal delay, which Tor cannot deliver. Internet telephony and video
conferencing services are simply unusable over metadata-protecting systems like Tor.
36.

As a result, although individuals can use security technologies to protect the contents of

their communications, there exist significant technical barriers that make it difficult, if not
impossible, to hide communications metadata, particularly for real-time communications
services like Internet telephony and video conferencing.
37.

Over the last three decades, and especially with the widespread adoption of mobile

phones in the past decade, our reliance on telecommunications has significantly increased.
Mobile phones are today ubiquitous, and their use necessarily requires reliance on a service
provider to transmit telephone calls, text messages, and other data to and fro. These
communications inevitably produce telephony metadata, which is created whenever a person
places a call. There is no practical way to prevent the creation of telephony metadata, or to erase
it after the fact. The only reliable way to avoid creating such metadata is to avoid telephonic
communication altogether.
Telephony Metadata Reveals Content
38.

Telephony metadata can be extremely revealing, both at the level of individual calls and,

especially, in the aggregate.

13

Case 1:13-cv-03994-WHP Document 27

39.

Filed 08/26/13 Page 14 of 35

Although this metadata might, on first impression, seem to be little more than

“information concerning the numbers dialed,” 24 analysis of telephony metadata often reveals
information that could traditionally only be obtained by examining the contents of
communications. That is, metadata is often a proxy for content.
40.

In the simplest example, certain telephone numbers are used for a single purpose, such

that any contact reveals basic and often sensitive information about the caller. Examples include
support hotlines for victims of domestic violence 25 and rape, 26 including a specific hotline for
rape victims in the armed services. 27 Similarly, numerous hotlines exist for people considering
suicide, 28 including specific services for first responders, 29 veterans, 30 and gay and lesbian
teenagers. 31 Hotlines exist for suffers of various forms of addiction, such as alcohol, 32 drugs, and
gambling. 33
24

Administration White Paper, Bulk Collection of Telephony Metadata Under Section 215 of
the USA Patriot Act 15 (Aug. 9, 2013), http://huff.to/1ey9ua5.
25

National Domestic Violence Hotline, The Hotline (last visited Aug. 22, 2013),
http://www.thehotline.org.
26

National Sexual Assault Hotline, RAINN: Rape, Abuse & Incest National Network (last
visited Aug. 22, 2013), http://www.rainn.org/get-help/national-sexual-assault-hotline.
27

About the Telephone Helpline, DOD Safe Helpline (last visited Aug. 22, 2013),
https://www.safehelpline.org/about-safe-helpline.
28

District of Columbia/Washington D.C. Suicide & Crisis Hotlines, National Suicide Hotlines
(last visited Aug. 22, 2013), http://www.suicidehotlines.com/distcolum.html.
29

Get Help Now! Contact us to Get Confidential Help via Phone or Email, Safe Call Now
(last visited Aug. 22, 2013), http://safecallnow.org.
30

About the Veterans Crisis Line, Veterans Crisis Line (last visited Aug. 22, 2013),
http://www.veteranscrisisline.net/About/AboutVeteransCrisisLine.aspx.
31

We Provide Crisis Intervention and Suicide Prevention for LGBTQ Youth, The Trevor
Project (last visited Aug. 22, 2013), thttp://www.thetrevorproject.org.
32

Alcohol Addiction Helpline, Alcohol Hotline (last visited Aug. 22, 2013),
http://www.alcoholhotline.com.
33

What is Problem Gambling?, National Council on Problem Gambling (last visited Aug. 22,
2013), http://bit.ly/cyosu.
14

Case 1:13-cv-03994-WHP Document 27

41.

Filed 08/26/13 Page 15 of 35

Similarly, inspectors general at practically every federal agency—including the NSA 34—

have hotlines through which misconduct, waste, and fraud can be reported, while numerous state
tax agencies have dedicated hotlines for reporting tax fraud. 35 Hotlines have also been
established to report hate crimes, 36 arson, 37 illegal firearms 38 and child abuse. 39 In all these cases,
the metadata alone conveys a great deal about the content of the call, even without any further
information.
42.

The phone records indicating that someone called a sexual assault hotline or a tax fraud

reporting hotline will of course not reveal the exact words that were spoken during those calls,
but phone records indicating a 30-minute call to one of these numbers will still reveal
information that virtually everyone would consider extremely private.
43.

In some cases, telephony metadata can reveal information that is even more sensitive than

the contents of the communication. In recent years, wireless telephone carriers have partnered
with non-profit organizations in order to permit wireless subscribers to donate to charities by
sending a text message from their telephones. These systems require the subscriber to send a
specific text message to a special number, which will then cause the wireless carrier to add that

34

Barton Gellman, NSA Statements to the Post, Wash. Post, Aug. 15, 2013,
http://wapo.st/15LliAB.
35

Report Tax Fraud – Tax Fraud Hotline, North Carolina Department of Revenue (last
visited Aug. 22, 2013), http://www.dor.state.nc.us/taxes/reportfraud.html.
36

Report Hate Crimes, LAMBDA GLBT Community Services (last visited Aug. 22, 2013),
http://www.lambda.org/hatecr2.htm.
37

ATF Hotlines – Arson Hotline, Bureau of Alcohol, Tobacco, Firearms and Explosives (last
visited Aug. 22, 2013), http://www.atf.gov/contact/hotlines/index.html.
38

ATF Hotlines – Report Illegal Firearms Activity, Bureau of Alcohol, Tobacco, Firearms
and Explosives (last visited Aug. 22, 2013), http://www.atf.gov/contact/hotlines/index.html.
39

Childhelp National Child Abuse Hotline, Childhelp (last visited Aug. 22, 2013),
http://www.childhelp.org/pages/hotline-home.

15

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 16 of 35

donation to the subscriber’s monthly telephone bill. For example, by sending the word HAITI to
90999, a wireless subscriber can donate $10 to the American Red Cross.
44.

Such text message donation services have proven to be extremely popular. Today,

wireless subscribers can use text messages to donate to churches, 40 to support breast cancer
research, 41 and to support reproductive services organizations like Planned Parenthood. 42
Similarly, after a policy change in 2012 by the Federal Election Commission, political candidates
like Barack Obama and Mitt Romney were able to raise money directly via text message. 43
45.

In all these cases, the most significant information—the recipient of the donation—is

captured in the metadata, while the content of the message itself is less important. The metadata
alone reveals the fact that the sender was donating money to their church, to Planned Parenthood,
or to a particular political campaign.
46.

Although it is difficult to summarize the sensitive information that telephony metadata

about a single person can reveal, suffice it to say that it can expose an extraordinary amount
about our habits and our associations. Calling patterns can reveal when we are awake and asleep;
our religion, if a person regularly makes no calls on the Sabbath, or makes a large number of
calls on Christmas Day; our work habits and our social aptitude; the number of friends we have;
and even our civil and political affiliations.

40

Several Ways to Give, The Simple Church (2013), http://bit.ly/1508Mgw; Other Ways to
Give, North Point Church (last visited Aug. 22, 2013), http://bit.ly/16S3IkO.
41

Donate by Text, Susan G. Komen for the Cure (last visited Aug. 22, 2013),
http://sgk.mn/19AjGP7.
42

Help Support a New Future for Illinois Women and Families, Planned Parenthood of
Illinois (last visited Aug. 22, 2013), http://bit.ly/1bXI2TX.
43

Dan Eggen, Text to ‘GIVE’ to Obama: President’s Campaign Launches Cellphone
Donation Drive, Wash. Post, Aug. 23, 2012, http://bit.ly/16ibjCZ.

16

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 17 of 35

Aggregated Telephony Metadata Is Even More Revealing
47.

When call metadata is aggregated and mined for information across time, it can be an

even richer repository of personal and associational details.
48.

Analysis of metadata on this scale can reveal the network of individuals with whom we

communicate—commonly called a social graph. By building a social graph that maps all of an
organization’s telephone calls over time, one could obtain a set of contacts that includes a
substantial portion of the group’s membership, donors, political supporters, confidential sources,
and so on. Analysis of the metadata belonging to these individual callers, by moving one “hop”
further out, could help to classify each one, eventually yielding a detailed breakdown of the
organization’s associational relationships.
49.

For instance, metadata can help identify our closest relationships. Two people in an

intimate relationship may regularly call each other, often late in the evening. If those calls
become less frequent or end altogether, metadata will tell us that the relationship has likely ended
as well—and it will tell us when a new relationship gets underway. More generally, someone
you speak to once a year is less likely to be a close friend than someone you talk to once a week.
50.

Even our relative power and social status can be determined by calling patterns. As The

Economist observed in 2010, “People at the top of the office or social pecking order often
receive quick callbacks, do not worry about calling other people late at night and tend to get
more calls at times when social events are most often organized (sic), such as Friday
afternoons.” 44

44

Mining Social Networks: Untangling the Social Web, Economist, Sep. 2, 2010,
http://econ.st/9iH1P7.

17

Case 1:13-cv-03994-WHP Document 27

51.

Filed 08/26/13 Page 18 of 35

At times, by placing multiple calls in context, metadata analysis can even reveal patterns

and sensitive information that would not be discoverable by intercepting the content of an
individual communication.
52.

Consider the following hypothetical example: A young woman calls her gynecologist;

then immediately calls her mother; then a man who, during the past few months, she had
repeatedly spoken to on the telephone after 11pm; followed by a call to a family planning center
that also offers abortions. A likely storyline emerges that would not be as evident by examining
the record of a single telephone call.
53.

Likewise, although metadata revealing a single telephone call to a bookie may suggest

that a surveillance target is placing a bet, analysis of metadata over time could reveal that the
target has a gambling problem, particularly if the call records also reveal a number of calls made
to payday loan services.
54.

With a database of telephony metadata reaching back five years, many of these kinds of

patterns will emerge once the collected phone records are subjected to even the most basic
analytic techniques.
55.

With an organization such as the ACLU, aggregated metadata can reveal sensitive

information about the internal workings of the organization and about its external associations
and affiliations. The ACLU’s metadata trail reflects its relationships with its clients, its
legislative contacts, its members, and the prospective whistleblowers who call the organization.
Second-order analysis of the telephony metadata of the ACLU’s contacts would then reveal even
greater details about each of those contacts. For example, if a government employee suddenly
begins contacting phone numbers associated with a number of news organizations and then the
ACLU and then, perhaps, a criminal defense lawyer, that person’s identity as a prospective

18

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 19 of 35

whistleblower could be surmised. Or, if the government studied the calling habits of the ACLU’s
members, it could assemble a detailed profile of the sorts of individuals who support the ACLU’s
mission.
56.

I understand from the plaintiffs that they sometimes represent individuals in so-called

“John Doe” lawsuits, where the individuals filing suit request anonymity—and are granted it by
the courts—because they are juveniles or because they wish to conceal sensitive medical or
psychiatric conditions. In such cases, analysis of aggregated metadata might reveal the
anonymous litigant. If, for example, the lawyers in the case have only a handful of contacts in
common other than mutual co-workers, and one or more of the lawyers generally call the same
one of those common contacts shortly before or after hearings or deadlines in the lawsuit, this
would imply the identity of the anonymous litigant. If the attorneys’ calling patterns suggest
more than one possible identity for the “John Doe,” metadata analysis of the candidate
individuals could verify the identity of the “John Doe,” by correlating facts about the individuals
with facts detailed in the lawsuit—for example, that he lives in a particular area (based on the
area code of his phone or those of the majority of his contacts), that he has a particular job (based
on calls made during work hours), that he has a particular medical condition (based on calls to
medical clinics or specialists), or that he holds particular religious or political views (based on
telephone donations, calls to political campaigns, or contact with religious organizations).
57.

Metadata analysis could even expose litigation strategies of the plaintiffs. Review of the

ACLU’s telephony metadata might reveal, for example, that lawyers of the organization
contacted, for example, an unusually high number of individuals registered as sex offenders in a
particular state; or a seemingly random sample of parents of students of color in a racially

19

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 20 of 35

segregated school district; or individuals associated with a protest movement in a particular city
or region.
58.

In short, aggregated telephony metadata allows the government to construct social graphs

and to study their evolution and communications patterns over days, weeks, months, or even
years. Metadata analysis can reveal the rise and fall of intimate relationships, the diagnosis of a
life-threatening disease, the telltale signs of a corporate merger or acquisition, the identity of a
prospective government whistleblower, the social dynamics of a group of associates, or even the
name of an anonymous litigant.
Mass Collection of Metadata and Data-Mining Across Many Individuals
59.

Advances in the area of “Big Data” over the past few decades have enabled researchers to

observe even deeper patterns by mining large pools of metadata that span many telephone
subscribers.
60.

Researchers have studied databases of call records to analyze the communications

reciprocity in relationships, 45 the differences in calling patterns between mobile and landline
subscribers, 46 and the social affinity and social groups of callers. 47
61.

Researchers have discovered that individuals have unique calling patterns, regardless of

which telephone they are using, 48 they have figured out how to predict the kind of device that is

45

Lauri Kovanen, Jari Saramaki & Kimmo Kaski, Reciprocity of Mobile Phone Calls,
Dynamics of Socio-Economic Systems (Feb. 3, 2010), http://arxiv.org/pdf/1002.0763.pdf.
46

Heath Hohwald, Enrique Frias-Martinez & Nuria Oliver, User Modeling for
Telecommunication Applications: Experiences and Practical Implications 8, (Data Mining and
User Modeling Group, Telefonica Research, 2013), http://bit.ly/1d7WkUU (“Interestingly,
Monday is the day with most calls for landline users, while Friday is the day with most calls for
mobile users. . . Mobile users spend less time on the phone than landline users.”).
47

Sara Motahari, Ole J. Mengshoel, Phyllis Reuther, Sandeep Appala, Luca Zoia & Jay Shah,
The Impact of Social Affinity on Phone Calling Patterns: Categorizing Social Ties from Call
Data Records, The 6th SNA-KDD Workshop (Aug. 12, 2012), http://b.gatech.edu/1d6i4RY.

20

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 21 of 35

making the calls (a telephone or a fax machine), 49 developed algorithms capable of predicting
whether the phone line is used by a business or for personal use, 50 identified callers by social
group (workers, commuters, and students) based on their calling patterns, 51 and even estimated
the personality traits of individual subscribers. 52
62.

The work of these researchers suggests that the power of metadata analysis and its

potential impact upon the privacy of individuals increases with the scale of the data collected and
analyzed. It is only through access to massive datasets that researchers have been able to identify
or infer new and previously private facts about the individuals whose calling records make up the
telephone databases. Just as multiple calls by the same person reveal more than a single call, so
too does a database containing calling data about millions of people reveal more information
about the individuals contained within it than a database with calling data about just one person.
As such, a universal database containing records about all Americans’ communications will
reveal vastly more information, including new observable facts not currently known to the

48

Corrina Cortes, Daryl Pregibon & Chris Volinsky, Communities of Interest, AT&T
Shannon Research Labs, http://www.research.att.com/~volinsky/papers/portugal.ps.
49

Haim Kaplan, Maria Strauss & Mario Szegedy, Just the Fax – Differentiating Voice and
Fax Phone Lines Using Call Billing Data, AT&T Labs, http://bit.ly/19Aa8Ua.
50

Corinna Cortes & Daryl Pregibon, Giga-Mining, AT&T Labs-Research,
http://bit.ly/153pMcI.
51

Richard A. Becker, Ramon Caceres, Karrie Hanson, Ji Meng Loh, Simon Urbanek,
Alexander Varshavsky & Chris Volinsky, Clustering Anonymized Mobile Call Detail Records to
Find Usage Groups, AT&T Labs-Research, http://soc.att.com/16jmKdz.
52

Rodrigo de Oliveira, Alexandros Karatzoglou, Pedro Concejero, Ana Armenta & Nuria
Oliver, Towards a Psychographic User Model from Mobile Phone Usage, CHI 2011 Work-inProgress (May 7–12, 2011), http://bit.ly/1f51mOy; see also Yves-Alexandre de Montjoye, Jordi
Quoidbach, Florent Robic & Alex (Sandy) Pentland, Predicting People Personality Using Novel
Mobile Phone-Based Metrics. Social Computing, Behavioral-Cultural Modeling and Prediction
(2013), http://bit.ly/1867vWU.

21

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 22 of 35

research commW1ity, because no researcher has access to the kind of dataset that the government
is presumed to have.
63 .

A common theme is seen in many of these examples of "big data" analysis of metadata.

The analyst uses metadata about many individuals to discover patterns of behavior that are
indicative of some attribute of an individual. The analyst can then apply these patterns to the
metadata of an individual user, to infer the likely attributes of that user. In this way, the effect of
collecting metadata about one individual is magnified when information is collected across the
whole population.
64.

The privacy impact of collecting all communications metadata about a single person for

long periods of time is qualitatively different than doing so over a period of days. Similarly, the
privacy impact of assembling the call records of every American is vastly greater than the impact
of collecting data about a single person or even groups of people. Mass collection not only
allows the government to learn information about more people, but it also enables the
government to learn new, previously private facts that it could not have learned simply by
collecting the information about a few, specific individuals.

Edward W. Felten
Dated: August 23 , 2013

22

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 23 of 35

EXHIBIT 1

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 24 of 35

Edward W. Felten
Professor of Computer Science and Public Affairs
Director, Center for Information Technology Policy
Princeton University
Sherrerd Hall, Room 302
Princeton NJ 08544
[phone]
[phone] fax
[e-mail]

Education
Ph.D. in Computer Science and Engineering, University of Washington, 1993.
Dissertation title: “Protocol Compilation: High-Performance Communication for
Parallel Programs.” Advisors: Edward D. Lazowska and John Zahorjan.
M.S. in Computer Science and Engineering, University of Washington, 1991.
B.S. in Physics, with Honors, California Institute of Technology, 1985.

Employment
Professor of Computer Science and Public Affairs, Princeton University, 2006-present.
Chief Technologist, U.S. Federal Trade Commission, 2011-2012.
Professor of Computer Science, Princeton University, 2003-2006.
Associate Professor of Computer Science, Princeton University, 1999-2003.
Assistant Professor of Computer Science, Princeton University, 1993-99.
Senior Computing Analyst, Caltech Concurrent Computing Project, California Institute
of Technology, 1986-1989.
Director, Center for Information Technology Policy, Princeton University, 2005-present.
Elysium Digital LLC and various law firms. Consulting and expert testimony in
technology litigation, 1998-present
U.S. Federal Trade Commission: consulting regarding spam policy and investigation,
2004, 2006.
U.S. Dept. of Justice, Antitrust Division: consulting and testimony in Microsoft antitrust
case, 1998-2002..
Electronic Frontier Foundation. Consulting in intellectual property / free speech lawsuits,
2001-2010.
Certus Ltd.: consultant in product design and analysis, 2000-2002.
Cigital Inc.: Technical Advisory Board member, 2000-2007.

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 25 of 35

Cloakware Ltd.: Technical Advisory Board member, 2000-2003.
Propel.com: Technical Advisory Board member, 2000-2002.
NetCertainty.com: Technical Advisory Board member, 1999-2002.
FullComm LLC: Scientific Advisory Board member, 1999-2001.
Sun Microsystems: Java Security Advisory Board member, 1997-2001.
Finjan Software: Technical Advisory Board member, 1997-2002.
International Creative Technologies: consultant in product design and analysis, 1997-98.
Bell Communications Research: consultant in computer security research, 1996-97.

Honors and Awards
National Academy of Engineering, 2013.
American Academy of Arts and Sciences, 2011
ACM Fellow, 2007.
EFF Pioneer Award, 2005.
Scientific American Fifty Award, 2003.
Alfred P. Sloan Fellowship, 1997.
Emerson Electric, E. Lawrence Keyes Faculty Advancement Award, Princeton
University School of Engineering, 1996.
NSF National Young Investigator award, 1994.
Outstanding Paper award, 1997 Symposium on Operating Systems Principles.
Best Paper award, 1995 ACM SIGMETRICS Conference.
AT&T Ph.D. Fellowship, 1991-93.
Mercury Seven Foundation Fellowship, 1991-93.

Research Interests
Information security. Privacy. Technology law and policy. Internet software.
Intellectual property policy. Using technology to improve government. Operating
systems. Interaction of security with programming languages and operating systems.
Distributed computing. Parallel computing architecture and software.

Professional Service
Professional Societies and Advisory Groups
ACM U.S. Public Policy Committee, Vice Chair, 2008-2010, 2012-present.
DARPA Privacy Panel, 2010-2012.
Transportation Security Administration, Secure Flight Privacy Working Group, 2005.
National Academies study committee on Air Force Information Science and Technology
Research, 2004-present.
Electronic Frontier Foundation, Advisory Board, 2004-2007.
ACM U.S. Public Policy Committee, 2004-present (Executive Committee, 2005-present)

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 26 of 35

ACM Advisory Committee on Security and Privacy, 2002-2003.
DARPA Information Science and Technology (ISAT) study group, 2002-2004.
Co-chair, ISAT study committee on “Reconciling Security with Privacy,” 2001-2002.
National Academy study committee on Foundations of Computer Science, 2001-2004.
Program Committees
World Wide Web Conference, 2006.
USENIX General Conference, 2004.
Workshop on Foundations of Computer Security, 2003.
ACM Workshop on Digital Rights Management, 2001.
ACM Conference on Computer and Communications Security, 2001.
ACM Conference on Electronic Commerce, 2001.
Workshop on Security and Privacy in Digital Rights Management, 2001.
Internet Society Symposium on Network and Distributed System Security, 2001.
IEEE Symposium on Security and Privacy, 2000.
USENIX Technical Conference, 2000.
USENIX Windows Systems Conference, 2000.
Internet Society Symposium on Network and Distributed System Security, 2000.
IEEE Symposium on Security and Privacy, 1998.
ACM Conference on Computer and Communications Security, 1998.
USENIX Security Symposium, 1998.
USENIX Technical Conference, 1998.
Symposium on Operating Systems Design and Implementation, 1996.
Boards
Electronic Frontier Foundation, Board of Directors, 2007-2010.
DARPA Information Science and Technology study board, 2001-2003.
Cigital Inc.: Technical Advisory Board.
Sun Microsystems, Java Security Advisory Council.
Cloakware Ltd.: Technical Advisory Board.
Propel.com: Technical Advisory Board.
Finjan Software: Technical Advisory Board.
Netcertainty: Technical Advisory Board.
FullComm LLC: Scientific Advisory Board.

University and Departmental Service
Committee on Online Courses, 2012-present
Director, Center for Information Technology Policy, 2005-present.
Committee on the Course of Study, 2009-present.
SEAS Strategic Planning, 2004.
Member, Executive Committee
Co-Chair, Interactions with Industry area.
Co-Chair, Engineering, Policy, and Society area.
Faculty Advisory Committee on Policy, 2002-present.
Council of the Princeton University Community, 2002-present (Executive Committee)
Faculty Advisory Committee on Athletics, 1998-2000.

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 27 of 35

Computer Science Academic Advisor, B.S.E. program, class of 1998 (approx. 25
students)
Faculty-Student Committee on Discipline, 1996-98.
Faculty-Student Committee on Discipline, Subcommittee on Sexual Assault and
Harrassment, 1996-98.

Students Advised
Ph.D. Advisees:
Harlan Yu (Ph.D. 2012). Dissertation: Designing Software to Shape Open Government
Policy.
Ariel J. Feldman (Ph.D. 2012). Dissertation: Privacy and Integrity in the Untrusted
Cloud.
Joseph A. Calandrino (Ph.D. 2012). Dissertation: Control of Sensitive Data in Systems
with Novel Functionality.
William B. Clarkson (Ph.D. 2012). Dissertation: Breaking Assumptions: Distinguishing
Between Seemingly Identical Items Using Cheap Sensors. Technical staff member at
Google.
Matthias Jacob (Ph.D. 2009). Technical staff member at Nokia.
J. Alex Halderman (Ph.D. 2009). Dissertation: Security Failures in Non-traditional
Computing Environments. Assistant Professor of Computer Science, University of
Michigan.
Shirley Gaw (Ph.D. 2009). Dissertation: Ideals and Reality: Adopting Secure
Technologies and Developing Secure Habits to Prevent Message Disclosure.
Technical staff member at Google.
Brent Waters (Ph.D. 2004). Dissertation: Security in a World of Ubiquitous Recording
Devices. Assistant Professor of Computer Science, University of Texas.
Robert A. Shillingsburg (Ph.D. 2004). Dissertation: Improving Distributed File Systems
using a Shared Logical Disk. Retired; previously a technical staff member at Google.
Michael Schneider (Ph.D. 2004). Dissertation: Network Defenses against Denial of
Service Attacks. Researcher, Supercomputing Research Center, Institute for Defense
Analyses.
Minwen Ji (Ph.D. 2001). Dissertation: Data Distribution for Dynamic Web Content.
Researcher, HP Labs.
Dirk Balfanz (Ph.D. 2000). Dissertation: Access Control for Ad Hoc Collaboration.
Technical staff member at Google.
Dan S. Wallach (Ph.D. 1998). Dissertation: A New Approach to Mobile Code Security.
Associate Professor of Computer Science, Rice University.
Significant Advisory Role:
Drew Dean (Ph.D. 1998). Advisor: Andrew Appel. Program Manager at DARPA.
Stefanos Damianakis (Ph.D. 1998). Advisor: Kai Li. President and CEO, Netrics, Inc.

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 28 of 35

Pei Cao (Ph.D. 1996). Advisor: Kai Li. Staff technologist at Facedbook.
Lujo Bauer (Ph.D. 2003). Advisor: Andrew Appel. Research Scientist, School of
Computer Science, Carnegie Mellon University.

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 29 of 35

Publications
Books and Book Chapters
[1] Enabling Innovation for Civic Engagement. David G. Robinson, Harlan Yu, and
Edward W. Felten. In Open Government, Daniel Lathrop and Laurel Ruma, eds.,
O’Reilly, 2010.
[2] Securing Java: Getting Down to Business with Mobile Code. Gary McGraw and
Edward W. Felten. John Wiley and Sons, New York 1999.
[3] Java Security: Web Browsers and Beyond. Drew Dean, Edward W. Felten, Dan S.
Wallach, and Dirk Balfanz. In "Internet Besieged: Countering Cyberspace
Scofflaws," Dorothy E. Denning and Peter J. Denning, eds. ACM Press, New York,
1997.
[4] Java Security: Hostile Applets, Holes and Antidotes. Gary McGraw and Edward
Felten. John Wiley and Sons, New York, 1996
[5] Dynamic Tree Searching. Steve W. Otto and Edward W. Felten. In "High
Performance Computing", Gary W. Sabot, ed., Addison Wesley, 1995.
Journal Articles
[6] Government Data and the Invisible Hand. David Robinson, Harlan Yu, William
Zeller, and Edward W. Felten. Yale Journal of Law and Technology, vol. 11, 2009.
[7] Mechanisms for Secure Modular Programming in Java. Lujo Bauer, Andrew W.
Appel, and Edward W. Felten. Software – Practice and Experience, 33:461-480,
2003.
[8] The Digital Millennium Copyright Act and its Legacy: A View from the Trenches.
Illinois Journal of Law, Technology and Policy, Fall 2002.
[9] The Security Architecture Formerly Known as Stack Inspection: A Security
Mechanism for Language-based Systems. Dan S. Wallach, Edward W. Felten, and
Andrew W. Appel. ACM Transactions on Software Engineering and Methodology,
9:4, October 2000.
[10] Statically Scanning Java Code: Finding Security Vulnerabilities. John Viega, Tom
Mutdosch, Gary McGraw, and Edward W. Felten. IEEE Software, 17(5), Sept./Oct.
2000.
[11] Client-Server Computing on the SHRIMP Multicomputer. Stefanos N. Damianakis,
Angelos Bilas, Cezary Dubnicki, and Edward W. Felten. IEEE Micro 17(1):8-18,
February 1997.
[12] Fast RPC on the SHRIMP Virtual Memory Mapped Network Interface. Angelos
Bilas and Edward W. Felten. IEEE Transactions on Parallel and Distributed
Computing, February 1997.

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 30 of 35

[13] Implementation and Performance of Integrated Application-Controlled File Caching,
Prefetching and Disk Scheduling. Pei Cao, Edward W. Felten, Anna R. Karlin, and
Kai Li. ACM Transactions on Computer Systems, Nov 1996.
[14] Virtual Memory Mapped Network Interface Designs. Matthias A. Blumrich, Cezary
Dubnicki, Edward W. Felten, Kai Li, and Malena Mesarina. IEEE Micro, 15(1):2128, February 1995.
Selected Symposium Articles
[15] Social Networking with Frientegrity: Privacy and Integrity with an Untrusted
Provider. Ariel J. Feldman, Aaron Blankstein, Michael J. Freedman, and Edward W.
Felten. Proc. USENIX Security Symposium, Aug. 2012.
[16] Bubble Trouble: Off-Line De-Anonymization of Bubble Forms. Joseph A.
Calandrino, William Clarkson, and Edward W. Felten. Proc. USENIX Security
Symposium, Aug. 2011
[17] You Might Also Like: Privacy Risks of Collaborative Filtering. Joseph A.
Calandrino, Ann Kilzer, Arvind Narayanan, Edward W. Felten, and Vitaly
Shmatikov. Proc. IEEE Symposium on Security and Privacy, May 2011.
[18] SPORC: Group Collaboration Using Untrusted Cloud Resources. Ariel J. Feldman,
William P. Zeller, Michael J. Freedman, and Edward W. Felten. Proc. Symposium
on Operating Systems Design and Implementation, 2010.
[19] SVC: Selector-Based View Composition for Web Frameworks. William Zeller and
Edward W. Felten. Proc. USENIX Conference on Web Application Development,
2010.
[20] Defeating Vanish with Low-Cost Sybil Attacks Against Large DHTs. Scott
Wolchok, Owen S. Hofmann, Nadia Heninger, Edward W. Felten, J. Alex
Halderman, Christopher J. Rossbach, Brent Waters, and Emmet Witchel. Proc. 17th
Network and Distributed System Security Symposium, 2010.
[21] Can DREs Provide Long-Lasting Security? The Case of Return-Oriented
Programming and the AVC Advantage. Stephen Checkoway, Ariel J. Feldman,
Brian Kantor, J. Alex Halderman, Edward W. Felten, and Hovav Shacham, Proc.
Electronic Voting Technology Workshop, 2009.
[22] Some Consequences of Paper Fingerprinting for Elections. Joseph A. Calandrino,
William Clarkson, and Edward W. Felten. Proc. Electronic Voting Technology
Workshop, 2009.
[23] Software Support for Software-Independent Auditing. Gabrielle A. Gianelli,
Jennifer D. King, Edward W. Felten, and William P. Zeller. Proc. Electronic Voting
Technology Workshop, 2009.
[24] Fingerprinting Blank Paper Using Commodity Scanners. William Clarkson, Tim
Weyrich, Adam Finkelstein, Nadia Heninger, J. Alex Halderman, and Edward W.
Felten. Proc. ACM Symposium on Security and Privacy, May 2009.

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 31 of 35

[25] Lest We Remember: Cold Boot Attacks on Encryption Keys. J. Alex Halderman,
Seth D. Schoen, Nadia Heninger, William Clarkson, William Paul, Joseph A.
Calandrino, Ariel J. Feldman, Jacob Appelbaum, and Edward W. Felten. Proc.
Usenix Security Symposium, 2008.
[26] In Defense of Pseudorandom Sample Selection. Joseph A. Calandrino, J. Alex
Halderman, and Edward W. Felten. Proc. Electronic Voting Technology Workshop,
2008.
[27] Security Analysis of the Diebold AccuVote-TS Voting Machine. Ariel J. Feldman,
J. Alex Halderman, and Edward W. Felten. Proc. Electronic Voting Technology
Workshop, 2007.
[28] Machine-Assisted Election Auditing. Joseph A. Calandrino, J. Alex Halderman, and
Edward W. Felten. Proc. Electronic Voting Technology Workshop, 2007.
[29] Lessons from the Sony CD DRM Episode. J. Alex Halderman and Edward W.
Felten. Proc. Usenix Security Symposium, 2006.
[30] A Convenient Method for Securely Managing Passwords. J. Alex Halderman, Brent
R. Waters, and Edward W. Felten. Proc. 14th World Wide Web Conference, 2005.
[31] New Client Puzzle Outsourcing Techniques for DoS Resistance. Brent R. Waters,
Ari Juels, J. Alex Halderman, and Edward W. Felten. ACM Conference on
Computer and Communications Security. November 2004.
[32] Privacy Management for Portable Recording Devices. J. Alex Halderman, Brent R.
Waters, and Edward W. Felten. 3rd Workshop on Privacy in Electronic Society.
November 2004.
[33] Receiver Anonymity via Incomparable Public Keys. Brent R. Waters, Edward W.
Felten, and Amit Sahai. ACM Conference on Computer and Communications
Security. November 2003.
[34] Attacking an Obfuscated Cipher by Injecting Faults. Matthias Jacob, Dan Boneh,
and Edward W. Felten. ACM Workshop on Digital Rights Management, November
2002.
[35] A General and Flexible Access-Control System for the Web. Lujo Bauer, Michael
A. Schneider, and Edward W. Felten. 11th USENIX Security Symposium, August
2002.
[36] Informed Consent in the Mozilla Browser: Implementing Value-Sensitive Design.
Batya Friedman, Daniel C. Howe, and Edward W. Felten. Hawaii International
Conference on System Sciences, January 2002. (Best Paper award, organizational
systems track.)
[37] Reading Between the Lines: Lessons from the SDMI Challenge. Scott A. Craver,
John P. McGregor, Min Wu, Bede Liu, Adam Stubblefield, Ben Swartzlander, Dan
S. Wallach, Drew Dean, and Edward W. Felten. USENIX Security Symposium,
August 2001.

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 32 of 35

[38] Cookies and Web Browser Design: Toward Realizing Informed Consent Online.
Lynette I. Millett, Batya Friedman, and Edward W. Felten. Proc. of CHI 2001
Conference on Human Factors in Computing Systems, April 2001.
[39] Timing Attacks on Web Privacy. Edward W. Felten and Michael A. Schneider. Proc.
of 7th ACM Conference on Computer and Communications Security, Nov. 2000.
[40] Archipelago: An Island-Based File System for Highly Available and Scalable
Internet Services. USENIX Windows Systems Symposium, August 2000.
[41] Proof-Carrying Authentication. Andrew W. Appel and Edward W. Felten. Proc. of
6th ACM Conference on Computer and Communications Security, Nov. 1999.
[42] An Empirical Study of the SHRIMP System. Matthias A. Blumrich, Richard D.
Alpert, Yuqun Chen, Douglas W. Clark, Stefanos, N. Damianakis, Cezary Dubnicki,
Edward W. Felten, Liviu Iftode, Margaret Martonosi, Robert A. Shillner, and Kai Li.
Proc. of 25th International Symposium on Computer Architecture, June 1998.
[43] Performance Measurements for Multithreaded Programs. Minwen Ji, Edward W.
Felten, and Kai Li. Proc. of 1998 SIGMETRICS Conference, June 1998.
[44] Understanding Java Stack Inspection. Dan S. Wallach and Edward W. Felten. Proc.
of 1998 IEEE Symposium on Security and Privacy, May 1998.
[45] Extensible Security Architectures for Java. Dan S. Wallach, Dirk Balfanz, Drew
Dean, and Edward W. Felten. Proc. of 16th ACM Symposium on Operating Systems
Principles, Oct. 1997. Outstanding Paper Award.
[46] Web Spoofing: An Internet Con Game. Edward W. Felten, Dirk Balfanz, Drew
Dean, and Dan S. Wallach. Proc. of 20th National Information Systems Security
Conference, Oct. 1997.
[47] Reducing Waiting Costs in User-Level Communication. Stefanos N. Damianakis,
Yuqun Chen, and Edward W. Felten. Proc. of 11th Intl. Parallel Processing
Symposium, April 1997.
[48] Stream Sockets on SHRIMP. Stefanos N. Damianakis, Cezary Dubnicki, and
Edward W. Felten. Proc. of 1st Intl. Workshop on Communication and Architectural
Support for Network-Based Parallel Computing, February 1997. (Proceedings
available as Lecture Notes in Computer Science #1199.)
[49] Early Experience with Message-Passing on the SHRIMP Multicomputer. Richard D.
Alpert, Angelos Bilas, Matthias A. Blumrich, Douglas W. Clark, Stefanos
Damianakis, Cezary Dubnicki, Edward W. Felten, Liviu Iftode, and Kai Li. Proc. of
23rd Intl. Symposium on Computer Architecture, 1996.
[50] A Trace-Driven Comparison of Algorithms for Parallel Prefetching and Caching.
Tracy Kimbrel, Andrew Tomkins, R. Hugo Patterson, Brian N. Bershad, Pei Cao,
Edward W. Felten, Garth A. Gibson, Anna R. Karlin, and Kai Li. Proc. of 1996
Symposium on Operating Systems Design and Implementation.
[51] Java Security: From HotJava to Netscape and Beyond. Drew Dean, Edward W.
Felten, and Dan S. Wallach. Proc. of 1996 IEEE Symposium on Security and
Privacy.

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 33 of 35

[52] Integrated Parallel Prefetching and Caching. Tracy Kimbrel, Pei Cao, Edward W.
Felten, Anna R. Karlin, and Kai Li. Proc. of 1996 SIGMETRICS Conference.
[53] Software Support for Virtual Memory-Mapped Communication. Cezary Dubnicki,
Liviu Iftode, Edward W. Felten, and Kai Li. Proc. of Intl. Parallel Processing
Symposium, April 1996.
[54] Protected, User-Level DMA for the SHRIMP Network Interface. Matthias A.
Blumrich, Cezary Dubnicki, Edward W. Felten, and Kai Li. Proc. of 2nd Intl.
Symposium on High-Performance Computer Architecture, Feb. 1996
[55] Improving Release-Consistent Shared Virtual Memory using Automatic Update .
Liviu Iftode, Cezary Dubnicki, Edward W. Felten, and Kai Li. Proc. of 2nd Intl.
Symposium on High-Performance Computer Architecture, Feb. 1996
[56] Synchronization for a Multi-Port Frame Buffer on a Mesh-Connected
Multicomputer. Bin Wei, Gordon Stoll, Douglas W. Clark, Edward W. Felten, and
Kai Li. Parallel Rendering Symposium, Oct. 1995.
[57] A Study of Integrated Prefetching and Caching Strategies. Pei Cao, Edward W.
Felten, Anna R. Karlin, and Kai Li. Proc. of 1995 ACM SIGMETRICS Conference.
Best Paper award.
[58] Evaluating Multi-Port Frame Buffer Designs for a Mesh-Connected Multicomputer.
Gordon Stoll, Bin Wei, Douglas W. Clark, Edward W. Felten, Kai Li, and Patrick
Hanrahan. Proc. of 22nd Intl. Symposium on Computer Architecture.
[59] Implementation and Performance of Application-Controlled File Caching. Pei Cao,
Edward W. Felten, and Kai Li. Proc. of 1st Symposium on Operating Systems
Design and Implementation, pages 165-178, November 1994.
[60] Application-Controlled File Caching Policies. Pei Cao, Edward W. Felten, and Kai
Li. Proc. of USENIX Summer 1994 Technical Conference, pages 171-182, 1994.
[61] Virtual Memory Mapped Network Interface for the SHRIMP Multicomputer.
Matthias A. Blumrich, Kai Li, Richard D. Alpert, Cezary Dubnicki, Edward W.
Felten, and Jonathan S. Sandberg. Proc. of Intl. Symposium on Computer
Architecture, 1994.
[62] Performance Issues in Non-Blocking Synchronization on Shared-Memory
Multiprocessors. Juan Alemany and Edward W. Felten. Proceedings of Symposium
on Principles of Distributed Computing, 1992.
[63] Improving the Performance of Message-Passing Applications by Multithreading.
Edward W. Felten and Dylan McNamee. Proceedings of Scalable High-Performance
Computing Conference (SHPCC), 1992.
[64] A Highly Parallel Chess Program. Edward W. Felten and Steve W. Otto. 1988
Conference on Fifth Generation Computer Systems.
Selected Other Publications

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 34 of 35

[65] Strangers in a Strange Land. Review of Blown to Bits: Your Life, Liberty, and
Happiness after the Digital Explosion, by Abelson, Ledeen, and Lewis. American
Scientist, 97:4. July/August 2009.
[66] Lest We Remember: Cold-Boot Attacks on Encryption Keys. J. Alex Halderman,
Seth D. Schoen, Nadia Heninger, William Clarkson, William Paul, Joseph A.
Calandrino, Ariel J. Feldman, Jacob Appelbaum, and Edward W. Felten.
Communications of the ACM, 52(5):91-98. May 2009.
[67] Security Analysis of the Diebold AccuVote-TS Voting Machine. Ariel J. Feldman,
J. Alex Halderman, and Edward W. Felten. Sept. 2006.
[68] Digital Rights Management, Spyware, and Security. Edward W. Felten and J. Alex
Halderman, IEEE Security and Privacy, Jan./Feb. 2006.
[69] Inside RISKS: DRM and Public Policy. Edward W. Felten. Communications of the
ACM, 48:7, July 2005.
[70] Understanding Trusted Computing: Will its Benefits Outweigh its Drawbacks?
Edward W. Felten. IEEE Security and Privacy, May 2003.
[71] A Skeptical View of DRM and Fair Use. Edward W. Felten. Communications of
the ACM 46(4):56-61, April 2003.
[72] Consumer Privacy and Government Technology Mandates in the Digital Media
Marketplace. Testimony before U.S. Senate Commerce Committee. September
2003.
[73] Secure, Private Proofs of Location. Brent R. Waters and Edward W. Felten.
Submitted for publication, 2003.
[74] An Efficient Heuristic for Defense Against Distributed Denial of Service Attacks
using Route-Based Distributed Packet Filtering. Michael A. Schneider and Edward
W. Felten. Submitted for publication, 2003.
[75] Written testimony to House Commerce Committee, Subcommittee on Courts, the
Internet, and Intellectual Property, oversight hearing on “Piracy of Intellectual
Property on Peer to Peer Networks.” September 2002.
[76] Written testimony to Senate Judiciary Committee hearings on “Competition,
Innovation, and Public Policy in the Digital Age: Is the Marketplace Working to
Protect Digital Creativity?” March 2002.
[77] Informed Consent Online: A Conceptual Model and Design Principles. Batya
Friedman, Edward W. Felten, and Lynette I. Millett. Technical Report 2000-12-2,
Dept. of Computer Science and Engineering, University of Washington, Dec. 2000.
[78] Mechanisms for Secure Modular Programming in Java. Lujo Bauer, Andrew W.
Appel, and Edward W. Felten. Technical Report CS-TR-603-99, Department of
Computer Science, Princeton University, July 1999.
[79] A Java Filter. Dirk Balfanz and Edward W. Felten. Technical Report 567-97, Dept.
of Computer Science, Princeton University, October 1997.

Case 1:13-cv-03994-WHP Document 27

Filed 08/26/13 Page 35 of 35

[80] Inside RISKS: Webware Security. Edward W. Felten. Communications of the ACM,
40(4):130, 1997.
[81] Simplifying Distributed File Systems Using a Shared Logical Disk.Robert A.
Shillner and Edward W. Felten. Princeton University technical report TR-524-96.
[82] Contention and Queueing in an Experimental Multicomputer: Analytical and
Simulation-based Results. Wenjia Fang, Edward W. Felten, and Margaret Martonosi.
Princeton University technical report TR-508-96.
[83] Design and Implementation of NX Message Passing Using SHRIMP Virtual
Memory Mapped Communication. Richard D. Alpert, Cezary Dubnicki, Edward W.
Felten, and Kai Li. Princeton University technical report TR-507-96.
[84] Protocol Compilation: High-Performance Communication for Parallel Programs.
Edward W. Felten. Ph.D. dissertation, Dept. of Computer Science and Engineering,
University of Washington, August 1993.
[85] Building Counting Networks from Larger Balancers. Edward W. Felten, Anthony
LaMarca, and Richard Ladner. Univ. of Washington technical report UW-CSE-9304-09.
[86] The Case for Application-Specific Communication Protocols. Edward W. Felten.
Univ. of Washington technical report TR-92-03-11.
[87] A Centralized Token-Based Algorithm for Distributed Mutual Exclusion. Edward W.
Felten and Michael Rabinovich. Univ. of Washington technical report TR-92-02-02.
[88] Issues in the Implementation of a Remote Memory Paging System. Edward W.
Felten and John Zahorjan. Univ. of Washington technical report TR-91-03-09.