9

speech uttered on the network, it rendered ineffective
many of the regulations by governments over the con-
tent of speech. States that previously controlled the ac-
cess by their citizens to certain political or cultural
speech found it hard to effect that control after the de-
ployment of the Internet. The Internet disabled content
regulation because its architecture had no facility for
distinguishing among different kinds of content. The
consequence was an expansion of free speech interna-
tionally. See, e.g., Peng Hwa Ang, Censorship and the
Internet: A Singapore Perspective,
< http://info.isoc.org/HMP/PAPER/132/abst.html>.

The Internet’s effect on privacy: Under the initial ar-
chitecture of the Internet, it was relatively difficult to
collect information about Internet users. The effect of
this architectural feature was that privacy was gener-
ally protected. With the release of Netscape Enterprise
Server 2.0, however, and the general rise in the use of
“cookie” technologies, it has become much easier to
monitor and track user behavior on the Internet. The
consequence of this change in technology has been to
render less effective protections for privacy. Code,
142-163.

The Internet’s effect on copyright. An early Internet
protocol — NNTP — enabled “USENET.” USENET
is a public messaging system, where users post mes-
sages to “newsgroups” that can then be read by mil-
lions of viewers across the world. While the number of
“newsgroups” in USENET is extremely large, among
the most popular are newsgroups that carry binary files
of erotic pictures. A large percentage of these pictures
are scanned images from Playboy and Penthouse
magazines, plainly posted without regard to rights of
the original copyright holders. The consequence of the
emergence of this protocol, enabling anyone anywhere
to become a “publisher,” has been the weakening of
copyright protection of commercial publishers. See,
e.g., Penthouse.com Moves Against Massive Copyright
Infringement
, Business Wire (May 2, 2000).

10

The Internet’s effect on limits on fraud. Email enables
the low cost delivery of apparently personal messages.
Unlike regular post, which can cost more than a dollar
per message, email is essentially free to the sender.
Programs for gathering email addresses from public
postings (such as USENET) have generated lists of
millions of email address which are often sold for less
than $100.00. The consequence is that a commercial
message can be delivered to millions at essentially no
cost. Such economics have led to a dramatic rise in the
quantity of “spam” sent by email. And as much of this
spam is fraudulent, this has meant a dramatic increase
in the burden of fraudulent commercial practices on
consumers. Lawrence Lessig & Paul Resnick, Zoning
Speech on the Internet: A Legal and Technical Model
,
98 Mich. L. Rev. 395, 427-29 (1999) (discussing the
control of “spam”) ; Federal Trade Commission,
Fighting Consumer Fraud: New Tools of the
Trade(April 1998).

The Internet’s effect on regulations protecting chil-
dren
. A relatively recent technology enabled by the
Internet has been web based chat services. These make
it easy for individuals to engage in real time conversa-
tion with others located around the world. Users of
these services enter a virtual room, and converse with
others in that room by typing messages that get associ-
ated with a user’s identity. Some of these services al-
low two individuals to choose to “leave” the “public”
room for a “private” room where conversation can oc-
cur without others viewing it. This technology has in-
creased the exposure of children to sexual abuse. Be-
cause adults can pretend to be children, they can lure
young children into conversations that then expose the
children to risks with the posing adult. Without this
technology, it would be relatively hard for the same
adults to engage in such conversations with kids
(thirty-five year old men roaming playgrounds are
usually easily noticed); with the technology, this
criminal activity is increased. See, e.g., Richard Green,
Pedophiles Prowling Streets of Cyberspace, The Post
and Courier (Charleston, SC) (June 18, 2000).

11

28. In each case, the technology of the Internet has had a conse-
quence for particular kinds of legal regulation or legal protec-
tion. Whether that regulation was a control of speech, the pro-
tection of privacy, the defense of copyright, the protection
against fraud, or the protection of children, the technology of
the Internet has weakened the effect of the legal regime that
existed prior to the Internet. The Internet changed the balance
of protection afforded by law, by enabling behaviors that
weaken the protections of a legal regime.

29. Governments have responded in different ways to this change
in the effectiveness of real space regulation. In some countries,
governments have banned or directly interfered with Internet
technologies. This interference has been effected either by re-
stricting use of the Internet, or by requiring that Internet activ-
ity pass through government controlled sites. See, e.g., Gere-
mie R. Barme and Sang Ye, The Great Firewall of China,
Wired 5.06 (June 1997). In other countries, the response has
been to target illegal behavior on the Internet, and to push for
new technologies that might help limit the effect of this illegal
behavior. See, for example, the policy of Hong Kong, at
< http://www.hkispa.org.hk/>.

30. The United States, and with few exceptions,4 has opted for the
second strategy. While acknowledging that behavior in cyber-
space is subject to legal regulation, in cases where the technol-
ogy of the Internet has made legal rules less effective, and
where the technology has had some legitimate use, the re-

4 I do not intend to overstate this characterization. The response of the United
States government has not been consistent, and the response of state govern-
ments has been much less careful. The United States government has banned,
for example, the export of certain encryption technologies, even though those
technologies, were they integrated into the network, would plainly facilitate
many legitimate uses of the network. See Bernstein v. U.S. Department of Jus-
tice, 176 F.3d 1132 (9th Cir 1999), withdrawn 1999 U.S. App. LEXIS 24324. So
too has the United States government regulated unauthorized access to computer
systems, even though an argument favoring access could well be made. Harold
Smith Reeves, Property in Cyberspace, 63 U. Chi. L. Rev. 761 (1996). At this
stage, however, the attitude of the Supreme Court is just beginning to work its
way through the legal system. It is therefore not surprising that there is some
inconsistency among different branches of the government.

12

sponse of the United States government has been to find ways
to better enforce the legal rule, rather than to disable, or ban, an
Internet technology. The general question has been how best to
counteract any illegitimate effect, given the range of tools for
regulating behavior: What mix of law and technology, in other
words, will best protect the legitimate state interests at stake,
without undermining the free speech and creativity that the In-
ternet makes possible?

31. This response of the U.S. government is not simply an abstract
question of public policy. In my view, the Supreme Court has
in effect mandated this response when legal regulations would
have a dramatic, or terminating, effect on an Internet technol-
ogy. That mandate flows from the Supreme Court’s opinion in
Reno v. ACLU. No doubt the legitimate state interest impli-
cated in Reno (protecting children from material “harmful to
minors”) is great. Indeed, in the view of many, this interest is
stronger than the interest in protecting copyright holders from
uncontrolled access to their speech. Yet before the regulation
of the CDA was permitted, the Court in Renorequired an
analysis of how the state regulation would burden the tech-
nologies of the Internet. Compare United States v. Playboy
Entertainment Group, Inc.,
120 S. Ct. 1878 (2000) (similar ap-
proach applied to cable regulation). And where alternative, less
drastic requirements would effect the same state end, the Court
invalidated a statute that did not use those less restrictive
means. At least during an initial period of Internet develop-
ment, the question is not just whether a regulation that affects
speech on the net advances a state end, but whether it advances
that end with the least possible burden on the technologies of
the Internet.

32. The same conclusion follows from the Supreme Court’s treat-
ment of VCR technology in the 1970s and 1980s. While copy-
right holders considered the VCR to be an illegal technol-
ogy—as MPA President Jack Valenti described the technology,
“the VCR is to the American film producer and the American

13

public as the Boston Strangler is to the woman alone,”5—the
Supreme Court was extraordinarily reluctant to use copyright
law to restrict VCR use or production. Sony Corporation v.
Universal City Studios,
464 U.S. 417 (1984). Instead, in evalu-
ating whether a technology could be banned because it could
be illegally used, the Court said that “the public interest in ac-
cess to that article of commerce is necessarily implicated.” Id.
at 440. So long as a technology is “merely … capable of sub-
stantial noninfringing uses,” id. at 442, the technology would
be protected.

IX.

Copyright and the Internet

33. Many have argued that no threat to legal protection on the In-
ternet is greater than the threat the Internet presents for copy-
right. For the same reason that cyberspace strengthens free
speech, it simultaneously renders the protection of copyrighted
material particularly vulnerable. In both cases, the original ar-
chitecture of cyberspace made it extremely hard to control ex-
change based on the content of the speech exchanged. In both
cases, controlling exchange based on the content of the speech
exchanged is the essence of the regulation. See the analysis in
Pamela Samuelson, The Copyright Grab, Wired 1.34 (1996).

34. Copyright holders have feared the threat posed by the Internet;
opponents of copyright have celebrated it. Both sides have
based their response upon a fact about the early architectures of
cyberspace: That the architecture enabled perfect copies of
digital content to be made for free, and that the early architec-
ture of the net made it possible for these perfect copies to be
distributed also essentially for free. These two features of the
initial architecture meant that digital content (unprotected by
technological protection measures) was extremely vulnerable
in cyberspace.

35.But in my view, based on a consideration of the interaction
between law and technology, both perspectives miss an im-

5Home Recording of Copyrighted Works, Hearings, Subcommittee of the Judi-
ciary Committee, 97th Cong., 2d Sess., Serial No. 97, pt 1, p. 8.

14

portant point. At the same time that cyberspace presents a
threat to copyrighted material, it also offers a much greater op-
portunity to control the access and use of copyrighted materi-
als. Code, 130-138. Technologies are being developed that can
radically increase the protection and control that copyright
holders have over their copyrighted material. These technolo-
gies will in turn, in my view, make it possible for copyright
holders to exercise more control over the use of copyrighted
material than they could in real space.

36.The best example of this technology is often referred to as
“trusted systems.” Code, 127-29. Xerox PARC’s Mark Stefik
is a leading researcher in this field. Stefik describes systems
where the copyright holders would not only be able to control
the duplication of copyrighted materials, but would also be able
to control the actual use of copyrighted materials. Trusted sys-
tems could control, for example, how often a text was ac-
cessed, or read. It could control whether the user had the ability
to cut parts of the text and paste them into other texts. It could
control whether the text could be printed, and how many times.
It could control whether the text could be shared. The technol-
ogy, in other words, would give the copyright holder a kind of
power over the user of copyrighted material that the copyright
holder has never before had. See Mark Stefik, Shifting the Pos-
sible: How Trusted Systems and Digital Property Rights Chal-
lenge Us to Rethink Digital Publishing,
12 Berkeley Tech. L.J.
137 (1997).

37. Trusted systems are not the only technology that would enable
copyright holders to better control their content. Technologies
for watermarking copyrighted material are proliferating; tech-
nologies for tagging copyrighted material are common; tech-
nologies for making copyrighted material uncopiable are
emerging. In each case, these technologies would enable the
copyright holders to regain a degree of control over their mate-
rial on the that the original architecture of the Internet re-
moved. There represent a change in the technologies of the In-
ternet that would strengthen the protection of law.

38. This increase in control implicates important copyright inter-
ests. As the Supreme Court noted in Sony, “the law has never
recognized an author’s right to absolute control of his work…”.
464 U.S., at 433 n.13. Yet these technologies could in effect

15

give an author “absolute control of his work.” This has led
many to worry that the architecture of cyberspace will create
too much, rather than too little, control over the distribution of
content. In other words, that the technology of cyberspace, in
combination with the protections of law, will produce greater
control over the use of copyrighted material than the balance
intended by the Copyright Act.

39. In this context, given (1) the ambiguous effect of Internet tech-
nologies on copyrighted speech, (2) the Supreme Court’s
strong protection for the free speech values embedded in the
original net, and (3) the Supreme Court’s rejection of “an
author’s right to absolute control of his work…” id., it is im-
portant, in my view, that society not act too quickly to skew le-
gal protection without a careful review of the actual effects of
any particular technology on the legal interests at stake.

40.This conclusion gains support from the free speech interests
that the Court relied upon in Reno. Copyright law, like the
regulation on “indecent speech” in Reno, is a regulation of
speech. Unlike the speech regulation in Reno, it may well be a
content neutral regulation of speech.6But whether content
neutral or content based, copyright law is still a regulation of
speech demanding heightened judicial scrutiny. Courts have
been instructed to move carefully in the context of speech
regulations that affect the Internet. If that caution was required
in the context of legislation aimed at protecting children, it
should be no less required in the context of legislation aimed at
protecting copyright holders. If there is a balance to be struck
judicially, it is to be struck only after an extensive hearing
about the effects of any such judicial regulation on net archi-
tecture.

IMAGE napd3.doc02.gif

6 A number of influential scholars have argued that copyright is a content based
regulation of speech. See, e.g., Mark A. Lemley & Eugene Volokh, Freedom of
Speech and Injunctions in Intellectual Property Cases,
48 Duke L.J. 147 (1998).
But see, e.g., Dr. Seuss Enters. v. Penguin Books USA, Inc., 109 F.3d 1394,
1403 n.11 (9th Cir. 1997) (treating it as content-neutral).

16

X.

Music and the Internet

41.The Internet has become an important mode for distributing
music. Through the use of compression technologies such as
the MP3 format, individuals are able to locate and download
relatively high fidelity copies of music from a practically end-
less array of sources. A large number of businesses have been
formed based upon this new mode of distribution. Among these
include eMusic.com and MP3.com.

42.These emerging businesses have a distinctive model for dis-
tributing content on the Internet that differs from the traditional
model for distributing music content in real space. Rather than
a few concentrated content distributors, the technology of the
Internet makes feasible many content providers, whose content
is identified and linked through sophisticated search and pref-
erence matching engines. As the cost of distribution would be
slight, and the cost of promotion and support thus less, this
model of distribution could well facilitate a greater diversity in
copyrighted content and musical sources. It could also, in the
view of many, facilitate a greater return to authors—the in-
tended beneficiaries of the Constitution’s Copyright Clause.

43. Some of the traditional record companies apparently resist this
change. They have responded to the possibilities of this differ-
ent model of distribution by selectively deploying the law to
weaken this emerging alternative to their method of distribut-
ing content. Rather than taking steps to protect their own con-
tent, while leaving other content to be freely exchanged in this
new mode of distribution, many of these companies have
sought to make illegal the technologies that support this new
mode of distribution. Their apparent aim is to use the law to fit
the Internet into their traditional business model.

44. It is perfectly legitimate, in my view, for these corporations to
lobby Congress to effect this end. But where the net benefit to
society of a new technology is uncertain, and where a balance
among competing interests must therefore be struck, it is be-
yond the judicial scope, as the Supreme Court indicated in
Sony, for courts to engage in a balancing of interests. Courts
are not to pick sides in a battle over modes of distribution for
music. Instead, if a new technology emerges that presents new

17

threats to an existing model of doing business, it has been the
practice of the Supreme Court to leave to Congress the task of
redrawing an appropriate balance.

45.This conclusion follows directly from the Supreme Court’s
treatment of the alleged threat presented by the VCR in Sony.
After considering a careful and extensive record provided by
District Court Judge Ferguson, 480 F. Supp. 429 (C.D. CA.
1979), the Court determined that so long as a technology was
“capable” of a “substantial noninfringing” use, it would not,
without more direction from Congress, ban the technology. The
issue was not whether the majority of actual uses of VCRs
were “fair uses” or not. The issue was whether the VCR was
“merely … capable of substantial noninfringing uses.” 464
U.S., at 442. As I explain below, there can be no doubt, in my
view, that the technologies at issue in this case are “capable of
substantial noninfringing uses,” and that these uses are them-
selves central to evolution of important technologies on the
Internet.

46. As the Court indicated in Sony, Congress frequently is called
upon to redraw the balance between the rights of copyright
holders and consumers in light of the emergence of new tech-
nologies. In light of the practice of home recording, for exam-
ple, Congress recently did draw a balance in the Audio Home
Recording Act. 17 U.S.C. § 1002. The express intent of Con-
gress in that act was to leave private, noncommercial home re-
cording unregulated by copyright law.7 If that is the proper in-
terpretation of the Act, and if the behavior of users of the Nap-
ster system fall within that category, then a great deal of the
activity using the Napster technology would be non-infringing
uses.8 If in light of the changes that the Internet has created,
Congress now wants to draw a different boundary — banning
the production and distribution of MP3 files, even for home or

IMAGE napd3.doc02.gif

7Congress did however regulate equipment to help protect copyright interests.
See Code, 46-47.

8 Regardless of how the Audio Home Recording Act is interpreted, a great many
uses of Napster are clearly non-infringing, including for example the distribution
of authorized works and the distribution of works in the public domain.

18

noncommercial use — Congress is free, subject to constitu-
tional limitations, to legislate in that direction. But the admoni-
tion of the Court in Sony is that courts alone do not make such
amendments to the copyright laws.

47. This conclusion is especially strong when, as in this case, the
very plaintiffs bringing the action have themselves facilitated
the distribution of copyrighted music without technologies for
enabling the control by the copyright holder. Sony Corporation,
for example, has pushed technologies that enable and encour-
age users to “rip” or copy the content of copyrighted CDs, and
to make that content available in a form that can be easily ex-
changed. They have done this even though there are many
technologies for enabling the distribution of music in a more
controlled format. (“LiquidAudio” is an example.) Thus, these
plaintiffs, on the one hand, are adding fuel to the fire of this
revolution in distribution, while, on the other hand, attempting
to shut down any avenue of distribution that they cannot con-
trol. This is an effort to extend the reach of their state-granted
monopoly rights. And while they are free to petition Congress
to extend their monopoly rights in this way, courts are not the
venue for such an extension.

XI.

Napster and Internet Architecture

48. Napster represents an evolution in Internet technologies. It is a
technology for facilitating more efficient searches for content
on peer systems, and for more efficiently transferring content
from peer systems. Neither of these two components alone rep-
resents a radical advance in Internet technology: the first is
simply a directory service, or “name space,” Erik Nilsson,
Napster: Popular Program Raises Devilish Issues, The
O’Reilly Network < http://www.oreillynet.com>; the second
simply uses the file transfer protocol (FTP) to transfer located
content. Reply Comment of Andrew Oram, Concerning the
Copyright Office’s Rulemaking on Exemptions from Prohibi-
tion on Circumvention of Technological Measures that Control
Access to Copyrighted Works,
March 31, 2000 (available at
< http://www.cpsr.org>). But Napster does combine these two
functionalities in an innovative way that results in making it

[made with GoClick]