Supreme Court Of The United States
No. 04-480
================================================================
In The
Supreme Court of the United States
----------------- ♦ -----------------
METRO-GOLDWYN-MAYER
STUDIOS, INC., et al.,
Petitioners,
v.
GROKSTER, LTD., et al.,
Respondents.
----------------- ♦ -----------------
On Writ Of Certiorari To The
United States Court Of Appeals
For The Ninth Circuit
----------------- ♦ -----------------
BRIEF AMICI CURIAE OF COMPUTER SCIENCE
PROFESSORS HAROLD ABELSON, THOMAS
ANDERSON, ANDREW W. APPEL, STEVEN M.
BELLOVIN, DAN BONEH, DAVID CLARK, DAVID J.
FARBER, JOAN FEIGENBAUM, EDWARD W.
FELTEN, ROBERT HARPER, M. FRANS
KAASHOEK, BRIAN KERNIGHAN, JENNIFER
REXFORD, JOHN C. REYNOLDS, AVIEL D. RUBIN,
EUGENE H. SPAFFORD AND DAVID S.
TOURETZKY SUGGESTING AFFIRMANCE
OF THE JUDGMENT
----------------- ♦ -----------------
VICTORIA K. HALL
JAMES S. TYRE*
LAW OFFICE OF
LAW OFFICES OF
VICTORIA K. HALL
JAMES S. TYRE
401 N. Washington St.
10736 Jefferson Blvd., #512
Suite
550
Culver City, CA 90230
Rockville MD 20850
(310) 839-4114
(301) 738-7677
*Counsel of Record
Counsel for Amici Curiae
February 28, 2005
================================================================
COCKLE LAW BRIEF PRINTING CO. (800) 225-6964
OR CALL COLLECT (402) 342-2831
i
TABLE OF CONTENTS
Page
TABLE OF AUTHORITIES ........................................ ii
INTERESTS OF AMICI CURIAE .............................. 1
SUMMARY OF ARGUMENT .....................................
4
ARGUMENT ............................................................... 5
I. NATURE OF THE INTERNET .......................
5
II. THE END-TO-END PRINCIPLE ....................
6
III.
THE DIFFICULTY OF DESIGNING DIS-
TRIBUTED NETWORKS ................................ 10
IV.
THE UNPROVEN EFFICACY OF CON-
TENT FILTERING TECHNOLOGIES ........... 14
V.
ANONYMITY .................................................. 18
CONCLUSION............................................................ 19
APPENDIX – BRIEF BIOGRAPHIES OF AMICI
CURIAE .......................................................................... 21
ii
TABLE OF AUTHORITIES
Page
CASES
A&M Records, Inc. v. Napster, Inc., 239 F.3d 1004
(9th Cir. 2001)................................................................. 17
A&M Records, Inc. v. Napster, Inc., 284 F.3d 1091
(9th Cir. 2002)........................................................... 17, 18
Sony Corp. of Am. v. Universal City Studios, Inc.,
464 U.S. 417 (1984) .................................................passim
Verizon Communications, Inc. v. Law Offices of
Curtis V. Trinko, LLP, 540 U.S. 398 (2004)..................... 3
MISCELLANEOUS
Bittorrent for torrent.linux.duke.edu, at http://torrent.
linux.duke.edu ................................................................ 12
John Borland, RIAA files 754 new file-swapping
suits, C|net news.com, Dec. 16, 2004, at http://
news.com.com/RIAA+files+754+new+file-swapping+
suits/2110-1027_3-5494259.html ................................... 19
John Borland, RIAA sues 717 file-swappers, C|net news.
com, Jan. 27, 2005, at http://news.com.com/RIAA+
sues+717+file-swappers/2110-1027_3-5553517.html .......... 19
David Cohen, New P2P Network Funded by U.S.
Government, New Scientist, Oct. 1, 2002, at http://
www.newscientist.com/article.ns?id=dn2861.................11
Steve Crocker, Request for Comments 1: Host Soft-
ware, at http://www.faqs.org/ftp/rfc/rfc1.txt (Apr. 7,
1969) ....................................................................................... 6
Daren Fonda, Downloading Hollywood, TIME, Feb.
14, 2005, at 43................................................................. 13
iii
TABLE OF AUTHORITIES – Continued
Page
IRIS: Infrastructure for Resilient Internet Systems,
at http://www.project-iris.net ..........................................11
Xeni Jardin, Hollywood Wants BitTorrent Dead, Wired
News, Dec. 14, 2004, at http://www.wired.com/news/
digiwood/0,1412,66034,00.html ............................................. 13
Kong is King.net|King Kong|Peter Jackson’s Pro-
duction Diary, at http://www.kongisking.net/kong
2005/proddiary................................................................ 12
Barry M. Leiner et al., A Brief History of the Inter-
net, at http://www.isoc.org/internet/history/brief.
shtml (last revised Dec. 10, 2003) ................................... 5
Marybeth Peters, Copyright Enters the Public
Domain, 51 J. Copyright Soc’y 701, 708 (2004) ............ 19
PunditGuy: Tsunami Videos, at http://www.punditguy.
com/2004/12/horror.html ................................................... 13
J.H. Saltzer et al., End-to-End Arguments in System
Design, 2 ACM Transactions on Computer Sys.
277-88 (Nov. 1984), available at http://mit.edu/
Saltzer/www/publications/endtoend/endtoend.pdf................ 7
Jonathan Zittrain & Ben Edelman, Empirical
Analysis of Internet Filtering in China, at http://
cyber.law.harvard.edu:8080/filtering/china................... 15
1
BRIEF AMICI CURIAE OF COMPUTER
SCIENCE PROFESSORS SUGGESTING
AFFIRMANCE OF THE JUDGMENT
These computer science professors, as amici curiae,
respectfully submit that the judgment below should be
affirmed.1
----------------- ♦ -----------------
INTERESTS OF AMICI CURIAE
As more fully described in the Appendix, amici are 17
computer science professors at nine major universities in
the United States.2 Each amicus respects the value of
intellectual property. All have published copyrighted works,
some hold patents, and some have seen their copyrighted
works made available without authorization on a peer-to-
peer (P2P) file-sharing network. None condone the unlaw-
ful use of file-sharing technology. Amici submit this brief
because amici are gravely concerned that the ability to
deploy or improve new technologies that can be used for
lawful and unlawful purposes will be severely constrained
if the Court scales back the protections inherent in the
1 Per Rule 37.6, amici state that no counsel for any party has
participated, in whole or in part, in writing this brief. The Distributed
Computing Industry Association is defraying the out-of-pocket cost of
printing this brief, but no person or entity other than amici or their
counsel has made any other monetary contribution for preparing or
submitting this brief. Counsel of record for amici is a Policy Fellow and
an Advisory Board member of the Electronic Frontier Foundation,
which is co-counsel for Respondent StreamCast Networks, Inc. Both
titles are unpaid and honorary designations, for work unrelated to this
case. The parties have consented to the filing of this brief.
2 Affiliations are listed only to identify the amici, whose views
expressed herein do not necessarily coincide with those of their
respective institutions.
2
“capable of substantial noninfringing uses” test of Sony
Corp. of Am. v. Universal City Studios, Inc., 464 U.S. 417
(1984) (Sony-Betamax).
Amici are technology innovators who have been
involved in major advances in Internet technology. David
Clark was one of the original designers of the Internet,
served as Chief Protocol Architect for the Internet in the
1980s, and has articulated the key design principles of the
Internet such as the end-to-end arguments, see Section III,
infra. Dr. Clark and his group created the first implemen-
tation of TCP/IP for the personal computer.3 Steven
Bellovin was one of the inventors of Usenet, an early and
highly decentralized network. David Farber was responsi-
ble for the development of the first distributed computer
system, was a principal in several major networking
efforts, and was the co-principal of the pioneering
NSF/DARPA-funded Gigabit Network Testbed Initiative.
Eugene Spafford developed the first generally available
intrusion detection system, and the first system security
scanner. He also was the first person to develop software
forensics techniques for cybercrime investigation. Other
amici have made and are making equally important
contributions to Internet technology.
Although some amici’s technology can be, has been
and will be used for unlawful purposes as well as for
lawful ones, amici have never before believed that their
work would be threatened by others’ unlawful use of their
technology. For 21 years, the “capable of substantial nonin-
fringing uses” test has protected those who developed new
3 TCP/IP (Transmission Control Protocol/Internet Protocol) is the
communication language that all computers use on the Internet.
3
technology capable of infringing and non-infringing uses –
technology that benefits all. While technology innovators
can guess at how a new technology will be used, only after
its release will they truly know its uses, either as deployed
initially or as modified.4 The fact is that Sony-Betamax has
been at the core of new technology development, because
technologists deliberately design for multiple uses, and
often cannot predict exactly what uses will be made of the
technology.
A year ago, this Court recognized the need to “ . . .
safeguard the incentive to innovate . . . .”5 Sony-Betamax’s
protection provides such an incentive, in academia as well
as in the private sector. If this Court should announce a
more restrictive rule, those who create the latest advances
in technology will halt or significantly scale back their
work, for fear of massive copyright infringement damages.
Such a rule will hinder technological progress, particularly
involving computers and the Internet. Almost all new
technology, whether coming from the private sector or
from academia, builds on that which has come before. The
best improvements, the best ideas, often derive from
seeing how a technology performs in actual use, not solely
4 Amici need look no further than Sony’s own Betamax machines.
They could record and play tapes of only one hour length, making
difficult “librarying” of longer programs and other uses that the studios
complained of at the time. See Sony-Betamax, 464 U.S. at 423. Other
companies developed the competing VHS standard, which allowed
playing and recording of longer tapes, and over time, that standard
drove the Betamax out of the market. It also allowed the videotape
rental and sale business to flourish, much to the benefit of the movie
studios’ profits.
5 Verizon Communications, Inc. v. Law Offices of Curtis V. Trinko,
LLP, 540 U.S. 398, 407 (2004) (for at least a limited time, innovators
can rely on monopoly power).
4
in a laboratory test bed. But if Petitioners have their way,
the ability to deploy or improve new technologies that can
be used for both lawful and unlawful purposes will be
constrained severely. This concern prompts amici to
submit this brief.6
----------------- ♦ -----------------
SUMMARY OF ARGUMENT
Amici write to call to the Court’s attention several
computer science issues raised by Petitioners and amici
who filed concurrent with Petitioners, and to correct
certain of their technical assertions. First, the United
States’ description of the Internet’s design is wrong. P2P
networks are not new developments in network design,
but rather the design on which the Internet itself is based.
Second, a P2P network design, where the work is done by
the end user’s machine, is preferable to a design which
forces work (such as filtering) to be done within the
network, because a P2P design can be robust and efficient.
Third, because of the difficulty in designing distributed
networks, advances in P2P network design – including
BitTorrent and Respondents’ software7 – are crucial to
developing the next generation of P2P networks, such as
6 Petitioners argue that the Court of Appeals misapplied the rule of
Sony-Betamax, rather than that the rule itself should be revisited.
However, amici apprehend no way to reconcile Petitioners’ arguments
with the “capable of substantial noninfringing uses” test of Sony-
Betamax.
7 Amici are aware that Grokster and the FastTrack protocol it uses
differ in many respects from Morpheus and the Gnutella protocol it
uses. Indeed those differences have value to academicians and tech-
nologists. However, for the purpose of this brief, amici need not
differentiate between the two.
5
the NSF-funded IRIS Project. Fourth, Petitioners’ asser-
tion that filtering software will work fails to consider that
users cannot be forced to install the filter, filtering soft-
ware is unproven or that users will find other ways to
defeat the filter. Finally, while Petitioners state that
infringers’ anonymity makes legal action difficult, the
truth is that Petitioners can obtain IP addresses easily
and have filed lawsuits against more than 8,400 alleged
infringers. Because Petitioners seek a remedy that will
hobble advances in technology, while they have other
means to obtain relief for infringement, amici ask the
Court to affirm the judgment below.
----------------- ♦ -----------------
ARGUMENT
I.
NATURE OF THE INTERNET.
First, amici address statements in the United States’
brief, see United States Br. at 2-3, that P2P design and file
sharing are recent aberrations. To the contrary, they have
been features of the Internet from its inception.8 Thus, any
liability rule applied to these technologies in general
applies in general to the Internet also.
A network system uses P2P design if it allows any
participant to act as a client, by requesting service from
another participant, and to act as a server, by providing
service in response to a client’s request. Any network that
treats its members as equals must use a P2P design.
8 See Barry M. Leiner et al., A Brief History of the Internet, at
http://www.isoc.org/internet/history/brief.shtml (last revised Dec. 10,
2003).
6
The very first Internet standards document,9 dated
April 7, 1969 and known as RFC 1, discusses the use of the
nascent network to connect any user to any remote com-
puter in what is now called a P2P fashion, and to transmit
files between computers via these connections. Indeed,
these are the only specific network building blocks (called
“primitives”) discussed in RFC 1.10 Development of P2P
interaction and file transfer has continued as the Internet
has grown. Accordingly, any rules that might be applied to
P2P technologies in general, or to file sharing systems in
general, necessarily would apply to the Internet in gen-
eral.
II.
THE END-TO-END PRINCIPLE.
Second, amici address assertions that checking for
infringement should be built into network design. On the
contrary, certain functionality (such as using filters)
should not be done at the network level. To order network
designers to add functionality to the network to avoid
liability is to force significant inefficiency into network
design. Because leaving out such functionality may repre-
sent good engineering design, no negative inference
regarding intent should be drawn if a designer chooses not
to add this functionality.
One of the most important principles of network design,
and one that underlies the Internet’s design, is the end-to-
end principle, which first appeared in a paper co-authored
9 Steve Crocker, Request for Comments 1: Host Software, at http://
www.faqs.org/ftp/rfc/rfc1.txt (Apr. 7, 1969).
10 Id. at 6.
7
by one of these amici, David Clark.11 The principle says
that most functions should be provided at the endpoints of
a network, rather than in the network itself. Because only
the endpoints know precisely what they want from the
network, the network cannot provide many functions
correctly and efficiently. Examples, using email and
teleconferencing, help explain the principle and why
adding functionality in networks poses problems.
Suppose we want to transmit an email message from
Alice to Bob, and we want to detect and correct any errors
that might creep into the message’s text between the time
Alice writes it and the time Bob reads it. Such errors could
creep in at several points: on Alice’s computer while the
message is waiting to be transmitted, on the network in
transit from Alice’s computer to Bob’s computer, or while
on Bob’s computer before he reads it. The best way to
provide appropriate error correction is to use an end-to-
end mechanism. If errors must be detected and corrected,
then a checksum12 can be used. Alice’s email program, at
one endpoint, adds a checksum to the message as soon as
she finishes writing it, and Bob’s email program, at the
other endpoint, verifies right before Bob reads the mes-
sage that the text is consistent with the checksum. If
errors are detected, Alice can retransmit the message.
11 J.H. Saltzer et al., End-to-End Arguments in System Design, 2
ACM Transactions on Computer Sys. 277-88 (Nov. 1984), available at
http://mit.edu/Saltzer/www/publications/endtoend/endtoend.pdf.
12 A checksum is a value computed from the content of a message
that can be used as a “signature” to verify its accuracy. Virtually any
change in a message’s length or content will alter the checksum,
allowing transmission errors to be readily detected.
8
If Alice and Bob want end-to-end protection, they need
to use an end-to-end mechanism as described above. Once
they do this, it would be redundant to add an error-
correction mechanism into the network itself. Worse yet,
adding error detection and correction would make the
network both more expensive, and less flexible for other
uses.
For example, in some applications, such as teleconfer-
encing, it is better to accept minor errors than to try to fix
them. A little snow on the screen is better than having the
image freeze entirely while the system tries to retransmit
the damaged image. A network that tried to fix errors
would be much worse for teleconferencing than one that
did not. Thus, adding functionality to a network can make
the network both more expensive and less useful.
Networks that are designed according to the end-to-
end principle are general-purpose, meaning that they can
be extended for a wide range of uses that were not antici-
pated at the time of their design. General-purpose net-
works derive their generality from the fact that they do
not try to understand the information that is flowing
through them. The telephone network, a familiar example
of a general-purpose network, is designed to carry any
voice (or voice-like audio signal) from Point A to Point B,
without trying to understand the meaning of the sound or
the content of the conversation, if indeed it is a conversa-
tion. Designing a network this way makes the network
easier to build. After all, it is much easier to transmit raw
voice sounds than to understand the content of speech.
But perhaps more important, because the telephone
network was designed to carry raw sounds, it could be
adapted later to a wide range of uses unforeseen by its
9
designers. Alexander Graham Bell probably did not foresee
faxes, answering machines, or voice mail. He could not
have predicted that some day vending machines would
make telephone calls when they ran out of candy bars. He
did not need to foresee these developments, because he
designed a general-purpose network that could support
whatever later uses that were found.
The Internet is the ultimate general-purpose network.
With its end-to-end design, the Internet serves a much
broader purpose than the telephone network, being able to
carry any type of digitized content, rather than only audio
signals. The general-purpose nature of the Internet is the
reason for its rapid evolution and its adoption by broad
segments of our society and economy.
In light of the astonishing usefulness of general-
purpose networks such as the telephone network and the
Internet, a designer’s decision not to include some function
in the network should not, in itself, be seen as evidence of
any particular intent. Amici have no knowledge of the
particular motives of Respondents, but caution against the
inference that a particular design decision, such as a
decision to include encryption or not to use filtering
technologies, necessarily represents bad faith. It may
simply represent good, conservative engineering. Here, the
United States seems to agree:
To the extent that petitioners’ argument concern-
ing vicarious liability could be construed as sug-
gesting the imposition of [ ] an obligation [on
Respondents to control their customers’ infring-
ing conduct], such a rule is neither desirable nor
supported by precedent. [ . . . ] The “right and
ability to supervise” element of vicarious liability
. . . has never, to our knowledge, been held to be
10
satisfied by the mere fact that the defendant
could restructure its relations or its product to
obtain such an ability. [ . . . ] The imposition of an
independent obligation to arrange one’s product
or relations in a way to permit the seller to re-
tain control would have the undesirable effect of
chilling technological innovation and constrain-
ing the product development options of develop-
ers of software and other digital technologies.
United States Br. at 19 n.3 (citations and quotes omitted).
III. THE DIFFICULTY OF DESIGNING DISTRIBUTED
NETWORKS.
Third, designing large-scale network systems poses
formidable challenges. To create the next generation of
networks, research scientists and software developers
constantly seek new and better network designs and
software. Due to the challenges in network design, and
numerous research questions in this area, amici respect-
fully urge the Court to be very cautious in addressing
liability rules so that such rules do not dictate the design
of such software, or order the redesign of systems that
function efficiently already.
Designing large-scale network systems is notoriously
difficult. Large networks must cope with vexing issues of
scale, reliability, robustness and security that simply do
not arise in smaller networks. Consequently, researchers
are looking more to P2P networks, which offer significant
advantages over client-server networks that have bottle-
necking problems when many users try to access a web
site, and can be easily taken down due to single points of
failure and denial of service attacks. When someone builds
successfully a large network, they teach valuable lessons
11
about the design of such networks. The networks created
by the users of Respondents’ software have certainly
taught such lessons, as have other P2P systems, which
designers use to advance research and product develop-
ment.
One beneficiary of such lessons is the National Sci-
ence Foundation-funded Infrastructure for Resilient
Internet Systems (IRIS) project.13 IRIS, co-led by one of
these amici, Frans Kaashoek, is a multi-institution col-
laboration, centered at MIT and U.C. Berkeley and funded
by a $12 million NSF grant.14 IRIS seeks to use a P2P
design strategy to support large-scale Internet services in
a manner more scalable, reliable, and secure than is
currently possible. Without Respondents’ success in
developing software that allows end users to participate in
a large scale P2P network, scientists on the IRIS Project
would have more difficulty in their research. Petitioners’
lawsuits against designers such as Respondents whose
software allows users to create a network threaten re-
search necessary to build better networks.
However, the chilling effect of lawsuits – actual or
threatened – is perhaps better illustrated by reference to
BitTorrent. Unlike IRIS, BitTorrent is a type of P2P
network software that is in widespread use already.
Developed by Bram Cohen, BitTorrent is an important
advance in large scale network technology because of its
13 IRIS: Infrastructure for Resilient Internet Systems, at http://
www.project-iris.net (last visited Feb. 6, 2005)
14 David Cohen, New P2P Network Funded by U.S. Government,
New Scientist, Oct. 1, 2002, at http://www.newscientist.com/article.ns?
id=dn2861.
12
usefulness in copying large files.15 BitTorrent allows a
large number of computers that have a file to share in
copying it to a person seeking it. Because the sharing is
simultaneous (each computer that has the file transfers a
portion of it at the same time as other computers that
have it) the transfer can avoid or lessen bottlenecking that
occurs if the entire file is copied from a single computer.
Further, as soon as the person has a portion of the file, her
computer shares in making it available to others who seek
it. In BitTorrent parlance, this is called swarming.
BitTorrent itself does not support file searching.
Consequently, a common way of determining whether a
file has been torrented (formatted so that it can be copied
using BitTorrent) is to look at a so-called tracker site: a
site that keeps track of torrented files, and allows one to
join in the swarm if one wants to copy a file. For example,
Red Hat, a major packager of Linux software, uses a
torrent tracker to save bandwidth in the distribution of its
software.16 As another example, Peter Jackson, the Pro-
ducer of the Lord of the Rings movie trilogy, now is produc-
ing a remake of King Kong. Jackson is keeping an online
production diary of the making of the film, that includes
both text and video. The video files being large, he is using
BitTorrent to share the work of distributing the files.17 As
still another, after the Tsunami, naturally there was great
15 Like any design, BitTorrent has strengths and weaknesses, and
amici can learn from both.
16 Duke University maintains the tracker for Red Hat’s Fedora
Core Linux software, Bittorrent for torrent.linux.duke.edu, at http://
torrent.linux.duke.edu (last visited Feb. 18, 2005).
17 Kong is King.net|King Kong|Peter Jackson’s Production Diary,
at http://www.kongisking.net/kong2005/proddiary (last visited Feb. 24,
2005).
13
interest in seeing the videos that had been taken on scene.
A number of trackers are available for those amateur
videos.18
In December 2004, the movie studios commenced a
series of legal actions against certain tracker sites – sites
that included tracks of the studios’ copyrighted works. At
the time, the studios seemed to appreciate the difference
between the technology and its developer on the one hand,
and unlawful uses of the technology on the other.19 The
MPAA anti-piracy chief acknowledged the existence of
legal torrent sites.20 However, by mid-February 2005, the
MPAA apparently had changed its thinking:
The industry is hoping that in a case scheduled
for next month, the U.S. Supreme Court will rule
against firms that produce file-sharing software,
such as Morpheus and Grokster. Neither Cohen
nor BitTorrent is named in the lawsuit, although
an MPAA spokesman says Cohen is under scru-
tiny for continuing to develop the software “and
making it easy to steal copyright material.”21
Clearly, BitTorrent not only is capable of substantial
noninfringing use, but in fact is used every day for sub-
stantial non-infringing purposes. A rule that would make a
developer like Cohen secondarily liable for copyright
18 See, e.g., PunditGuy: Tsunami Videos, at http://www.punditguy.
com/2004/12/horror.html (last visited Feb. 24, 2005).
19 See, e.g., Xeni Jardin, Hollywood Wants BitTorrent Dead, Wired
News, Dec. 14, 2004, at http://www.wired.com/news/digiwood/0,1412,66034,00.
html.
20 Id.
21 Daren Fonda, Downloading Hollywood, TIME, Feb. 14, 2005, at
43.
14
infringement, merely because his software can be and is
used for infringing purposes would also cripple advances
in large-scale network design. Such a rule surely would
not safeguard the incentive to innovate.
IV. THE UNPROVEN EFFICACY OF CONTENT
FILTERING TECHNOLOGIES.
Petitioners,
and
some
amici22 assert that Respondents
should use various content filtering technologies, in an
attempt to prevent infringement on the networks created
by their users. Contrary to such assertions, the efficacy of
these technologies is far from established. Amici wish to
bring three things to the Court’s attention.
First, the suggested filtering strategy would require
filtering software to be installed on users’ computers. Even
assuming that Respondents have the right and ability to
deliver such software to end users, there can be no way to
ensure that software updates are installed, and stay
installed. End users ultimately have control over which
software is on their computers. If an end user does not
want a software update, there is no way to make her take
it. Indeed, a computer user who exercises proper security
precautions should not allow any third party to install or
upgrade software on her machine.
22 United States Br. at 26; Am. Fed’n of Musicians of the U.S. &
Can. et al. Br. at 19; Audible Magic Corp. et al. Br.; Bridgemar Servs.
Ltd. Br.; Hollaar Br. at 20; Kids First Coalition et al. Br. at 8, 13;
Macrovision Corp. Br. at 8, 11; Napster, LLC et al. Br. at 14; Nat’l Acad.
of Recording Arts & Scis. et al. Br. at 18-21; Nat’l Ass’n of Recording
Merchandisers Br. at 7 n.5, 19 n.21; Office of the Comm’r of Baseball et
al. Br. at 14; Snocap, Inc. Br.; Utah et al. Br. at 20 & n.13, 24-26; Video
Software Dealers Ass’n Br. at 16.
15
Second, it is important to recognize that the filtering
technologies in question have not been subjected to any
significant public testing or scrutiny. No demonstration
has shown that these technologies would be effective in
distinguishing infringing from noninfringing files if
deployed in conjunction with software like Respondents’.
No demonstration has shown that these technologies
would scale to the extent necessary to be deployed success-
fully on large networks created by users of Respondents’
software. While some amici who have filtering products
they want to sell, have filed briefs in this case, some of
their products are not even available for sale yet, let alone
been subject to testing. Filter sellers, understandably,
have high hopes for their products. These hopes should not
be mistaken for evidence.
Third,
experience
shows
that users respond to filter-
ing and blocking technologies by devising methods to
defeat the filter. For example, the Chinese government
runs a filter, known colloquially as the “Great Firewall of
China,” that tries to restrict access from within China to
certain Internet material of which the Chinese govern-
ment disapproves – possibly including, at times, this
Court’s web site.23 Chinese citizens, with the help of
outsiders, have found many ways to defeat this firewall to
read the forbidden material.
To predict the effect of filtering technologies, a static
analysis, which assumes deployment of the technologies
but ignores the likely responses of noncompliant users, is
23 Jonathan Zittrain & Ben Edelman, Empirical Analysis of Internet
Filtering in China, at http://cyber.law.harvard.edu:8080/filtering/china
(last visited Feb. 18, 2005).
16
not sufficient. Instead, an accurate prediction of filter
software effectiveness must use a dynamic analysis, which
considers users’ responses.
If filters were introduced as suggested by Petitioners,
and assuming – despite lack of evidence – that the filters
were deployable and would work exactly as Petitioners
hope, noncompliant users would still have several methods
for defeating them. For example, a user could encrypt files,
to hide their contents, before submitting the files to
Respondents’ software; other users, on receiving the files,
would decrypt them to recover the original contents.
Respondents’ software (and any filter incorporated into it)
would then see the files only in encrypted form, and so
would be unable to distinguish infringing files from
noninfringing files. Suitable encryption tools are widely
available.
Filter designers might respond by trying to detect and
block encrypted files. That task is more difficult than it
sounds, and besides, noncompliant users could respond by
manipulating the encrypted files to look like innocuous
content, using standard information-hiding technology.
Each move by the filter designers would elicit a counter-
move from the noncompliant users.
Because users would respond to filtering technology
advances, the introduction of such technologies would not
solve the infringement problem but would kick off an
open-ended arms race between the filter designers and
noncompliant users. To enter this arms race would be to
take on recurring costs (in money and lost technical flexibil-
ity), in exchange for benefits that are at best uncertain.
Further, turning secondary liability on this inevitable
dynamic would embroil the courts in a continuing process
17
of evaluating and choosing new technologies intended to
respond to these countermeasures.
The Napster case illustrates these points. After
Napster I,24 the Court of Appeals remanded the case to the
district court for further proceedings. The district court
then ordered Napster to filter file names of copyrighted
works noticed by plaintiffs. A&M Records, Inc. v. Napster,
Inc., 284 F.3d 1091, 1096 (9th Cir. 2002) (Napster II).
When users responded by making alterations in file names
or the spelling of titles or artists’ names, Napster was
required to identify and filter those variations. Id. When
that did not work to the satisfaction of the district court,
Napster installed audio fingerprinting technology, technol-
ogy which relies on the “fingerprint” of the copyrighted
work, and thus is not dependent on the file name or
spelling adjustments. Id. at 1097.
Napster succeeded in “prevent[ing] sharing of much of
plaintiffs’ noticed copyrighted works.” Id. at 1096. How-
ever, that was not good enough. The district court de-
manded zero error tolerance for plaintiffs’ properly noticed
works. Id. at 1098. When Napster could not achieve
perfection, the court ordered Napster to shut down, and
the Court of Appeals affirmed. Id. at 1099.
Though
Petitioners
and
amici cite often to Napster I,
the near total absence of Napster II in their briefs shows a
reluctance to acknowledge that the filter software they
speak of is mostly “vaporware” – non-existent. Some amici
offer technology similar to that which failed to achieve
zero error tolerance in Napster II. Some make grand
24 A&M Records, Inc. v. Napster, Inc., 239 F.3d 1004 (9th Cir. 2001).
18
claims, but none guarantee or offer proof that they can
achieve anything close to the zero error tolerance required
in Napster II.
V. ANONYMITY.
Petitioners and some amici assert incorrectly that
users of P2P networks are anonymous and hence need not
fear enforcement.25 The truth is, Respondents’ software
does little to prevent its users from being identified.
Generally, providing anonymous, reliable communications
among strangers is a difficult task. No demonstration has
shown that a practical file-sharing system can provide
such anonymity.
Respondents’ software transfers files directly from the
users who have them to the users who request them.
Because of this, anyone who downloads a file learns the
IP address of the person who is providing that file.26
Anyone can connect to these networks and download files,
in order to learn the IP addresses of users who offer them.
The IP addresses can be used later to identify those who
provide copyrighted files.27 Petitioners (or their agents) and
other copyright owners have used this method to sue more
than 8,400 users of various P2P networks for direct
25 Pet. Motion Picture Studio & Recording Co. Br. at 3-4; Am. Soc’y
of Composers, Authors & Publishers et al. Br. at 9-10; Defenders of
Prop. Rights Br. at 4; Kids First Coalition et al. Br. at 17-19; Office of
Comm’r of Baseball et al. Br. at 4, 8-9.
26 An IP address is a numeric value that identifies a particular
“location” on the Internet.
27 A user’s IP address may change from time to time, but the user’s
Internet Service Provider will keep records that can be used to deter-
mine who was using a particular IP Address at a particular time.
19
infringement.28 Thus, arguments that anonymity is a
barrier to enforcement are false. As stated recently by the
Register of Copyrights, The Honorable Marybeth Peters:
Technology, however, makes [detection of in-
fringement] much more possible – an individual’s
activity on peer-to-peer networks can be moni-
tored and logged by the same computers that
make the reproduction and distribution possible.
This gives copyright owners a possibility of en-
forcement that they did not have before.
Marybeth Peters, Copyright Enters the Public Domain, 51
J. Copyright Soc’y 701, 708 (2004).
----------------- ♦ -----------------
CONCLUSION
Amici urge the Court not to be lured into abandoning
the Sony-Betamax “capable of substantial noninfringing
uses” test, which has protected advances in technology so
well. Abandoning this test will chill future development of
Internet technologies. Petitioners claim otherwise, but
Petitioners seem not to fully appreciate how new technolo-
gies are developed. From a purely technical standpoint,
Respondents’ products are not so very different from
mainstream Internet technologies. Attempts to regulate or
redesign large-scale network systems carry larger risks,
and offer benefits much less certain than Petitioners
28 See John Borland, RIAA files 754 new file-swapping suits, C|net
news.com, Dec. 16, 2004, at http://news.com.com/RIAA+files+754+new+
file-swapping+suits/2110-1027_3-5494259.html (7,706 lawsuits filed); John
Borland, RIAA sues 717 file-swappers, C|net news.com, Jan. 27, 2005, at
http://news.com.com/RIAA+sues+717+file-swappers/2110-1027_3-5553517.html
(717 more lawsuits).
20
would have the Court believe. Amici ask the Court to take
care to protect the scientists and engineers who are
developing tomorrow’s technology. Amici ask the Court to
preserve the Sony-Betamax “capable of substantial nonin-
fringing uses” test.
Respectfully
submitted,
VICTORIA K. HALL
JAMES S. TYRE*
LAW OFFICE OF
LAW OFFICES OF
VICTORIA K. HALL
JAMES S. TYRE
401 N. Washington St.
10736 Jefferson Blvd., #512
Suite
550
Culver City, CA 90230
Rockville MD 20850
(310) 839-4114
(301) 738-7677
*Counsel of Record
Counsel for Amici Curiae
February 28, 2005
21
APPENDIX –
BRIEF BIOGRAPHIES OF AMICI CURIAE
Harold
(Hal)
Abelson29 is Class of 1922 Professor of
Electrical Engineering and Computer Science at MIT and
a Fellow of the Institute of Electrical and Electronics
Engineers (IEEE). He is the winner of the 1995 Taylor L.
Booth Education Award given by IEEE Computer Society,
cited for his continued contributions to the teaching of
introductory computer science. At MIT, Abelson is co-
director of the MIT-Microsoft Research Alliance in Educa-
tional Technology and co-chair of the MIT Council on
Educational Technology. Together with his colleague
Gerald Sussman, Abelson developed MIT’s introductory
computer science subject, “Structure and Interpretation of
Computer Programs,” which has had a world-wide impact
on university computer-science education. Dr. Abelson
teaches a course in collaboration with Harvard Law
School, which deals with technical and policy issues
relating to the Internet, including copyright management,
content control, and privacy.
Thomas
Anderson30 is Professor in the Department of
Computer Science and Engineering at the University of
Washington. His research concerns the principles underly-
ing the construction of secure, reliable and efficient large
scale networks and distributed systems. He serves as Chair
of the Steering Committee of the Planetlab Consortium, an
29 Hal Abelson, Department of Electrical Engineering and Com-
puter Science, Massachusetts Institute of Technology, at http://www.swiss.
ai.mit.edu/~hal/hal.html (last visited Feb. 6, 2005).
30 Tom Anderson, Department of Computer Science and Engineer-
ing, University of Washington, at http://www.cs.washington.edu/homes/
tom (last visited Feb. 6, 2005).
22
association of over one hundred universities and corpora-
tions devoted to developing the next generation of Internet
technologies.31 Planetlab operates a network of over five
hundred machines, spread over five continents, that can
be used by researchers to test and deploy new planetary
scale applications. An author of over 70 research papers,
Anderson has won the National Science Foundation (NSF)
Presidential Faculty Fellowship and the Sloan Research
Fellowship.
Andrew
W.
Appel32 is a Professor of Computer Science
at Princeton University. He does research in computer
security, virus prevention, programming languages, and
compilers. He is a Fellow of the Association for Computing
Machinery (ACM) and served for several years as Editor in
Chief of ACM Transactions on Programming Languages
and Systems. Dr. Appel served as a primary technical
expert for nine non-settling States in the Microsoft anti-
trust trial, New York v. Microsoft.
Steven M. Bellovin33 is a Professor of Computer
Science at Columbia University. He joined the faculty
recently after many years at Bell Labs and AT&T Labs
Research. He is an AT&T Fellow and a member of the
National Academy of Engineering. Dr. Bellovin is the co-
author of Firewalls and Internet Security: Repelling the
31 PlanetLab Consortium, at http://www.planet-lab.org (last visited
Feb. 6, 2005).
32 Andrew W. Appel, Department of Computer Science, Princeton
University, at http://www.cs.princeton.edu/~appel (last visited Feb. 6,
2005).
33 Steven M. Bellovin, Department of Computer Science, Columbia
University, at http://www.cs.columbia.edu/~smb (last visited Feb. 6,
2005).
23
Wily Hacker (2d ed. 2003) (with amicus Aviel Rubin, and
Bill Cheswick) [hereinafter Firewalls and Internet Secu-
rity], and holds several patents on cryptographic and
network protocols. He has served on many National
Research Council (NRC) study committees, and is a
member of the Department of Homeland Security’s Science
and Technology Advisory Committee. He has been a
member of the Internet Architecture Board and co-director
of the Security Area of the Internet Engineering Task
Force.
Dan
Boneh34 is a Professor of both Electrical Engineer-
ing and Computer Science at Stanford University. He
heads the applied crypto group at the Computer Science
department. Dr. Boneh’s research focuses on applications
of cryptography to computer security. He is the author of
over 70 technical publications. His work includes digital
copyright protection, e-mail security, security for handheld
devices and web servers and cryptanalysis. He is a recipi-
ent of the Packard Award, the Alfred P. Sloan Award, and
the Terman Award.
David Clark35 is a Senior Research Scientist at MIT.
More recent activities than those stated in the brief
include extensions to the Internet to support real-time
traffic, pricing and related economic issues, and policy
issues surrounding the Internet, such as broadband local
loop deployment. His current research looks at re-definition
34 Dan Boneh, Department of Computer Science and Electrical
Engineering, Stanford University, at http://theory.stanford.edu/~dabo
(last visited Feb. 15, 2005).
35 David Clark, Computer Science and Artificial Intelligence
Laboratory, Massachusetts Institute of Technology, at http://www.lcs.
mit.edu/people/bioprint.php3?PeopleID=81 (last visited Feb. 6, 2005).
24
of the architectural underpinnings of the Internet, and the
relation of technology and architecture to economic,
societal and policy considerations. He is past chairman of
the NRC’s Computer Science and Telecommunications
Board and a member of the National Academy of Engi-
neering. Dr. Clark is a Fellow of the IEEE and of the ACM,
and has won numerous awards for his work.
David J. Farber36 is the Distinguished Career Profes-
sor of Computer Science and Public Policy at the School of
Computer Science at Carnegie Mellon University with
secondary appointments at the Heinz School and the
Engineering and Public Policy Department of the College
of Engineering. In 2003, he retired from the University of
Pennsylvania where he held the Alfred Fitler Moore Chair
of Telecommunications with appointments in the Engi-
neering School and the Wharton School. From 2000-01, he
served as Chief Technologist for the Federal Communica-
tions Commission. Prior to his appointment to the FCC, he
served on the U.S. Presidential Advisory Committee on
High Performance Computing and Communication,
Information Technology, and the Next Generation Internet
[hereinafter U.S. Presidential Advisory Committee on
Information Technology]. He is a Fellow of the ACM and
the IEEE, and serves on the Board of Directors of the
Electronic Frontier Foundation. He is a member of the
Markle Foundation Task Force on National Security in the
Information Age.
36 David J. Farber, School of Computer Science, Carnegie Mellon
University, at http://www.epp.cmu.edu/people/bios/farber.htm (last visited
Feb. 6, 2005).
25
Joan Feigenbaum37 is a Professor in the Computer
Science Department at Yale University. Prior to starting at
Yale in 2000, she worked for AT&T, where she participated
broadly in the company’s Information-Sciences research
agenda, e.g., by creating a research group in Algorithms
and Distributed Data, of which she was manager in 1998-
99. Dr. Feigenbaum’s research interests include Internet
algorithms, computational complexity, security and pri-
vacy, and digital copyright. While at Yale, she has been a
principal in several high-profile activities, including the
NSF-funded PORTIA (Privacy, Obligations, and Rights in
Technologies of Information Assessment) Project and the
Office of Naval Research-funded SPYCE (Stanford-Penn-
Yale-Cornell Experiment) Project. Her current and recent
professional service activities include Editor-in-Chief for
the Journal of Cryptology, Program Chair for the 2002
ACM Workshop on Digital Rights Management, and
Program Co-Chair for the 2004 ACM Conference on
Electronic Commerce. Dr. Feigenbaum is a Fellow of the
ACM.
Edward W. Felten38 is a Professor of Computer Science
at Princeton University. He is also affiliated with the
Program in Science, Technology, and Environmental Policy,
in the Woodrow Wilson School of Public and International
Affairs, at Princeton. His research interests include com-
puter security, Internet software, and information technology
37 Joan Feigenbaum, Computer Science Department, Yale Univer-
sity, at http://cs-www.cs.yale.edu/homes/jf/home.html (last visited Feb.
6, 2005).
38 Edward W. Felten, Department of Computer Science, Princeton
University, at http://www.cs.princeton.edu/~felten (last visited Feb. 6,
2005).
26
policy. He is widely known for his research on anti-copying
technologies. He has served in an advisory capacity to the
U.S. Departments of Defense, Justice, and Homeland
Security, and has testified before the Senate Commerce
Committee regarding digital copyright policy. He was the
primary computer science expert witness for the United
States in United States v. Microsoft. Dr. Felten is a mem-
ber of the Advisory Board of the Electronic Frontier
Foundation.
Robert
Harper39 is a Professor of Computer Science at
Carnegie Mellon University. His research is on program-
ming language design and implementation. He is a princi-
pal co-designer of the Standard ML programming
language and a co-inventor of the LF Logical Framework.
He is Associate Editor for Programming Languages of the
Journal of the ACM, the premier academic journal in
computer science.
M.
Frans
Kaashoek40 is a Professor in MIT’s Electrical
Engineering and Computer Science Department and a
member of the Computer Science and Artificial Intelli-
gence Laboratory, where he co-leads the parallel and
distributed operating systems group. Dr. Kaashoek’s
principal field of interest is designing and building com-
puter systems. His past work includes the exokernel
operating system, the Click modular router, the Resilient
Overlay Network, the self-certifying file system, and the
39 Robert Harper, School of Computer Science, Carnegie Mellon
University, at http://www-2.cs.cmu.edu/~rwh (last visited Feb. 15,
2005).
40 M. Frans Kaashoek, Department of Electrical Engineering and
Computer Science, Massachusetts Institute of Technology, at http://
www.pdos.lcs.mit.edu/~kaashoek (last visited Feb. 6, 2005).
27
Chord protocol, a robust, scalable protocol to locate infor-
mation in P2P systems. His current focus is the IRIS
project. Dr. Kaashoek is the recipient of several awards,
including the inaugural ACM Special Interest Group on
Operating Systems’ Mark Weiser award for demonstrating
creativity and innovation in operating systems research.
Brian Kernighan41 is a Professor in the Computer
Science Department at Princeton University. Previously he
was head of the Computing Structures Research Depart-
ment at Bell Labs, where he did research in programming
languages, software tools, and user interfaces. He is the
co-author of a number of widely-used computer books and
programs, and is a member of the National Academy of
Engineering.
Jennifer Rexford42 is a Professor in the Computer
Science Department at Princeton University. Prior to
February 2005, she spent eight years at AT&T Labs
Research. Her research on network measurement, traffic
engineering, and router configuration has led to several
network-management tools that are in daily use in AT&T’s
IP backbone network. She is co-author of Web Protocols
and Practice: HTTP/1.1, Networking Protocols, Caching,
and Traffic Measurement (2001). She serves as the chair of
ACM Special Interest Group on Data Communications,
and is a member of the ACM Council, the Computing
41 Brian Kernighan, Department of Computer Science, Princeton
University, at http://www.cs.princeton.edu/~bwk (last visited Feb. 6,
2005).
42 Jennifer Rexford, Department of Computer Science, Princeton
University, at http://www.cs.princeton.edu/~jrex (last visited Feb. 6,
2005).
28
Research Association’s Board of Directors, and DARPA’s
Information Science and Technology Study Group.
John
C.
Reynolds43 is Professor of Computer Science at
Carnegie Mellon University. He is a Fellow of the ACM
and a recipient of the Lifetime Achievement Award from
the ACM Special Interest Group on Programming Lan-
guages. He is also a member of International Federation of
Information Processing Working Group 2.3 on Program-
ming Methodology. His research centers on the design of
languages for programming and the specification of
programs, programming methodology, and
methods for
proving that programs meet their specifications.
Aviel
D.
Rubin44 is a Professor of Computer Science
and the Technical Director of the Information Security
Institute at Johns Hopkins University. Prior to joining
Johns Hopkins, Dr. Rubin was a scientist at AT&T Labs
Research. Dr. Rubin is author of several books including
Firewalls and Internet Security (co-authored with Bill
Cheswick and amicus Steven M. Bellovin), White-Hat
Security Arsenal (2001), and Web Security Sourcebook
(1997) (co-authored with Dan Geer and Marcus Ranum).
He is Associate Editor of IEEE Transactions on Software
Engineering, Associate Editor of ACM Transactions on
Internet Technology, Associate Editor of IEEE Security &
Privacy, and an Advisory Board member of Springer’s
Information Security and Cryptography Book Series. Dr.
43 John C. Reynolds, School of Computer Science, Carnegie Mellon
University, at http://www-2.cs.cmu.edu/~jcr (last visited Feb. 6, 2005).
44 Avi Rubin, Department of Computer Science, Johns Hopkins
University, at http://www.cs.jhu.edu/~rubin (last visited Feb. 6, 2005).
29
Rubin serves on DARPA’s Information Science and Tech-
nology Study Group.
Eugene
H.
Spafford45 is a Professor of both Computer
Sciences and of Electrical and Computer Engineering at
Purdue University with courtesy appointments as Profes-
sor of Communication and as Professor of Philosophy. He
is the founder and executive director of CERIAS, the
Purdue Center for Education and Research in Information
Assurance and Security, a national center of excellence
and the nation’s foremost academic center in this field. Dr.
Spafford is a Fellow of the ACM, the IEEE, and the Ameri-
can Association for the Advancement of Science, and he
was the year 2000 recipient of the NIST/NSA National
Computer Software Security Award. He has been named to
the Information Systems Security Association Hall of
Fame, and has been awarded the William Hugh Murray
medal from the National Colloquium for Information
Systems Security Education for his contributions to
research and education in information security. He is a
recipient of the Air Force medal for Meritorious Civilian
Service, is a recipient of the IEEE Computer Society’s
Taylor Booth medal, and of the ACM Special Interest
Group on Computers and Society’s “Making a Difference”
award. He currently serves on the Computing Research
Association’s Board of Directors, and on the U.S. Presiden-
tial Advisory Committee on Information Technology, as
well as many corporate advisory boards. Dr. Spafford has
co-authored or edited 5 books, over 100 technical articles,
holds 2 patents, and is the author of several software
45 Eugene H. Spafford, Department of Computer Science, Purdue
University, at http://www.cerias.purdue.edu/homes/spaf/narrate.html
(last visited Feb. 18, 2005).
30
packages that have been the basis for commercial prod-
ucts.
David S. Touretzky46 is a Research Professor of Com-
puter Science at Carnegie Mellon University, and co-
director of the graduate training program of the Center for
the Neural Basis of Cognition. He has lectured on artificial
intelligence, neural networks, and computational neuro-
science in formats ranging from half-day industry semi-
nars to semester-long graduate courses. Dr. Touretzky is
also the author of a popular textbook on the Lisp pro-
gramming language. He serves on the board of directors of
the NIPS (Neural Information Processing Systems) Foun-
dation, and is a member of the ACM and the American
Association for Artificial Intelligence.
46 David S. Touretzky, Computer Science Department, Carnegie
Mellon University, at http://www-2.cs.cmu.edu/~dst (last visited Feb. 6,
2005).
