September 9, 2003
Committee on the Judiciary
United States Senate
September 9, 2003
Pornography, Technology, and Process: Problems and Solutions on Peer-to-Peer Networks
Statement of the Honorable Marybeth Peters,
Register of Copyrights
Mr. Chairman, Senator Leahy, Members of the Committee, good afternoon. It is always a pleasure to appear before this Committee and I thank you for inviting me to present the views of the Copyright Office today at this very timely hearing. As you were among the leaders in drafting and enacting the Digital Millennium Copyright Act ("DMCA"), I know that these issues are important to you, as they are to me.
In 1999, a young man named Shawn Fanning developed a use of the Internet that allowed people to identify and copy music files from other people's computers. As you know, this model popularized peer-to-peer technology and a company called Napster tried to turn it into a profit-making business. Napster became phenomenally popular in a remarkably short period of time, boasting millions of registered users the very next year. But it quickly became clear that Napster was being used extensively (by millions of users) for the purpose of copying and distributing an unprecedented number of copyrighted works, primarily sound recordings of musical works.
That was the scene when you held a hearing on July 11, 2000, Mr. Chairman, entitled "Music on the Internet: Is There an Upside to Downloading?" At that hearing, Mr. Hank Barry, then the CEO of Napster, stated "It is my firm belief that the consumers who use Napster are not committing copyright violations." We did not agree with that assessment, and we were heartened when the Ninth Circuit found that "Napster users infringe at least two of the copyright holders' exclusive rights: the rights of reproduction...and distribution." Napster was unable to find a way to continue operations and faded away.
The void left by Napster's departure was filled by other businesses utilizing peer-to-peer technology, such as Aimster, Grokster, and Kazaa. While some of these applications can be differentiated from Napster in terms of their internal technical operation, they still follow the same basic peer-to-peer model as Napster and it is apparent that an overwhelming number of their customers are using it for the same purpose as they and others had used Napster - copying and distributing copyrighted works. By now it is well-settled that those users are infringing copyright. Notwithstanding that, there are still some who contend that such uses are not infringing.
Mr. Chairman, make no mistake. The law is unambiguous. Using peer-to-peer networks to copy or distribute copyrighted works without permission is infringement and copyright owners have every right to invoke the power of the courts to combat such activity. Every court that has addressed the issue has agreed that this activity is infringement. It can also be a crime and the perpetrators of such a crime are subject to fines and jail time.
Some have tried to rationalize or justify their illegal behavior by attacking the victim with allegations of inflated profits or unfair dealings with recording artists on the part of the recording industry. These diversionary tactics do not alter the fundamental fact that they are trying to defend illegal activity that takes place on peer-to-peer networks. For those who do not have sympathy for the recording industry, there are other victims as well. Since Napster, subsequent versions of peer-to-peer networks permit infringement of the works of other copyright owners, large and small, from motion picture studios to independent photographers and needlepoint designers. With broadband connections becoming more and more widespread, it is increasingly more common that the larger files containing full-length motion pictures are copied back and forth. This problem is not shrinking; it is not static; it is growing.
There are some who argue that copyright infringement on peer-to-peer systems is not truly harmful to copyright owners and may even help them generate new interest in their products. The law leaves that judgment to the copyright owner and it ought not be usurped by self-interested third parties who desire to use the copyright owner's work.
II. Copyright Liability of Peer-to-Peer Proprietors
Copyright law has long recognized that those who aid and abet copyright infringement are no less culpable than the direct infringers themselves. There are two types of this secondary liability. Contributory infringement occurs when "[o]ne who, with knowledge of the infringing activity, induces, causes, or materially contributes to the infringing conduct of another." For purposes of this test, knowledge can be either actual or constructive - that is, having reason to know. Vicarious liability occurs when one "has the right and ability to supervise the infringing activity and also has a direct financial interest in such activities."
Both of these concepts were brought to bear in the case against Napster. The Ninth Circuit agreed with the District Court that Napster had actual knowledge of the infringements it was facilitating from, for example, notices from aggrieved copyright owners. There was little question but that Napster provided a material contribution in the form of "the site and facilities" for infringement. Thus, Napster was determined to be a contributory infringer.
The Ninth Circuit also considered whether Napster was vicariously liable. It had no difficulty agreeing with the District Court that the infringing material on its network was a "draw" for customers, thus providing a direct financial benefit from the infringing activity. The Ninth Circuit also agreed with the District Court that Napster had the ability to police its system, and thus that it had the right and ability to supervise its users' conduct. Accordingly, Napster was found to be vicariously liable as well.
Thus it was that many felt reassured that the Ninth Circuit had confirmed that copyright law provides an effective and efficient way in which to address the massive infringements that can and do occur on peer-to-peer networks. Unfortunately, the Napster decision was not the final word on the matter.
Earlier this year, the Central District of California surprised many when it held that Grokster and Kazaa are not liable as secondary copyright infringers. This decision departed from long-established precedent. For example, the court held that in order to establish contributory liability, it must be shown that "a defendant has actual - not merely constructive - knowledge of the infringement at a time during which the defendant materially contributes to that infringement." Were such a standard to be adopted it would eviscerate the doctrine of contributory infringement as it would be almost impossible to meet. It would encourage the kind of sophistry we have seen from the proprietors of some peer-to-peer applications: a denial of knowledge of infringements by their customers in the face of clear and uncontested evidence that such infringement is occurring on a mind-boggling scale. Mr. Chairman, these are people whose business plan is dependant upon massive copyright infringement and any application of the law that allows them to escape liability for lack of knowledge of those same infringements is inherently flawed.
Not only was the Kazaa decision wrong on the law, it has serious policy consequences as well. The historical doctrines of secondary liability have served copyright owners, courts, and the public well - they provide copyright owners with the ability to obtain relief against the root cause of a series of infringements without costly, inefficient, and burdensome suits against numerous individuals. Without a viable doctrine of contributory liability, this option is severely curtailed and may present the copyright owner with the unenviable choice of either accepting unremedied infringements or filing numerous suits against the individual direct infringers.
If today's hearing leaves the Committee with the impression that the law is in flux with regard to the liability of proprietors of peer-to-peer technology, that is because it is. On one side is the Napster decision of the Ninth Circuit and the Aimster decision of the Seventh Circuit, both finding liability, albeit through different paths of analysis. On the other side is the Kazaa decision of the Central District of California, finding no liability for Kazaa and Grokster. Hanging over all of these cases is the Supreme Court's decision in Sony. It is perhaps a commentary on that opinion that almost twenty years later, we still have such uncertainty that three courts seem to interpret and apply it in three different ways. I believe that the correct application of the doctrines of secondary liability and the Sony case should produce findings of liability for the proprietors of Kazaa and Grokster as well as Napster and Aimster. If the case law evolves so as to compel the opposite result, I believe that Sony should be revisited either by the Supreme Court or by Congress.
III. Suits Against Individuals
Unless and until the Kazaa decision is overruled, copyright owners are faced with the unenviable choice to which I referred earlier. They can either resign themselves to unremedied infringements on a previously unimaginable scale, or they can file infringement actions against individual peer-to-peer users. The recording industry has chosen the latter route.
While copyright owners have expressed regret that they have felt compelled to take this step, they need offer no apologies. As I have already said, people who use peer-to-peer technology for the unauthorized reproduction or distribution of copyrighted works are breaking the law. Surprisingly, many people do not appear to realize this. I have long advocated more public education about copyright. In a perfect world, this could be done in classrooms and with billboards. But ours is not a perfect world, and public education can also be accomplished through enforcement of copyright.
The threats of litigation and even the publicity about the subpoenas obtained by the RIAA have made clear to everyone that the so-called "file-sharing" of copyrighted works is not an innocent activity without legal consequences. As a result, it is becoming more and more difficult for people engaged in such activity to claim that they did not know what they were doing is against the law. Of course, for some users of peer-to-peer technology, even knowledge that what they are doing is illegal will not be a sufficient disincentive to engage in such conduct. But whether or not these infringers know or care that it is against the law, the knowledge that such conduct may lead to expensive and burdensome litigation and a potentially large judgment should have a healthy deterrent effect. While we would like to think that everyone obeys the law simply because it is the law and out of a sense of obligation, we also know that laws without penalties may be widely ignored. For many people, the best form of education about copyright in the internet world is the threat of litigation. In short, if you break the law, you should be prepared to accept the consequences. Copyright owners have every right to enforce their rights in court, whether they are taking action against providers of peer-to-peer services designed to profit from copyright infringement or against the persons engaging in individual acts of infringement using such services.
IV. Statutory History and Interpretation of Subsection 512(h)
It is common sense that in order to be able to take action against the users of peer-to-peer networks, the copyright owner must know who those users are. Congress foresaw this need and addressed it by including in the DMCA a process by which copyright owners can learn basic identifying information about alleged infringers from their internet service providers ("ISPs").
The DMCA began as an effort to implement the 1996 WIPO Internet treaties. Neither those treaties nor any other international instrument directly address the potential secondary liability of ISPs. However, as the treaty implementing legislation moved forward in Congress, representatives of ISPs demanded that the legislation also limit their liability under such circumstances. Congress heeded this call and provided the ISPs with a huge benefit - virtually no liability for qualifying ISPs. This was balanced by a carefully developed set of obligations in the DMCA. Among those balancing obligations was the requirement that ISPs "expeditiously" respond to subpoenas to provide identifying information about subscribers accused of copyright infringement so that the controversy could be settled in court.
At the time the DMCA was drafted, at least one representative of ISPs assured this Committee that ISPs desired a solution whereby "service providers and content owners...work as a partnership...." It was asserted by that same representative that "[l]iability for copyright infringement should fall where it belongs, on the Web site operators, on those who create an infringing work or on those who reproduce it or perform it with actual knowledge of the infringement...." The ability of copyright owners to utilize subsection 512(h) is a critical part of that partnership as is copyright owners' ability to impose liability against those who infringe copyright. It is regrettable that at least one major ISP now rejects the compromise and the balance of the DMCA.
Some now claim that the subpoena power of subsection 512(h) is inapplicable to the activity described in subsection 512(a). As the United States District Court for the District of Columbia recently held, the plain language of subsection 512(h) demonstrates that this interpretation is not correct. I agree with the court's analysis.
Subsection 512(h) instructs service providers to expeditiously respond to a subpoena. The definition of "service provider" in section 512(k) always includes service providers which qualify for the safe harbor in section 512(a). The court reasoned that this demonstrates Congress' intent to apply the subpoena power to "all service providers, regardless of the functions a service provider may perform under the four categories set out in subsections (a) through (d)."
It has also been argued that the subpoena power applies only to subsection 512(c) because subsection 512(h)(2)(A) requires a copyright owner to supply "a copy of a notification described in subsection (c)(3)(A)". However, as the District Court pointed out, subsection 512(h) "is written without limitation or restriction as to its application." It does not require that a notice be delivered. Had Congress wished to limit the application of the subpoena power, it would have simply said so in the law. It did not.
The statutory text confirms the policy of compromise behind subsection 512 -- that copyright owners and ISPs work together to remedy infringement. Limiting the subsection 512(h) subpoena provisions as some have proposed would remove an important tool that parties need to remedy infringement efficiently in the peer-to-peer context.
When it enacted the DMCA, Congress did not carve out an exception from subsection 512(h) for transitory digital network communications, the activity covered by subsection 512(a). Service providers which engage in that activity received the benefits and burdens of the same bargain that service providers engaged in the other activity covered by section 512 received. In exchange for a powerful limitation on liability, they undertook some obligations, including the obligation to identify alleged infringers when served with a subsection 512(h) subpoena. When you enacted section 512, you made the right choice. There is no reason for the courts or Congress to have second thoughts about that decision.
I understand that the majority if not all of the 512(h) subpoenas that have been sought, have been sought in the United States District Court for the District of Columbia. Apparently this has necessitated the clerk of that court assigning additional staff to handle the workload. I do not take a position as to whether it is appropriate for a copyright owner to go to a single district court for subpoenas to service providers located outside that district. However, I am sympathetic to concerns about efficiency of the courts and fairness to ISPs located elsewhere in the country. There would certainly be advantages to the filing of these subpoena requests in the districts in which the ISPs are located.
V. Constitutional Challenges to Subsection 512(h)
The United States has intervened in the Verizon-RIAA litigation to defend the constitutionality of the DMCA. The Copyright Office has assisted the Justice Department in this effort and we firmly believe that subsection 512(h) is appropriate and constitutional. Although I am not an expert on constitutional law and I am not here to represent the Department of Justice, I will briefly summarize the arguments the United States made in its brief to the District Court.
The claim that subsection 512(h) violates the case and controversy requirement of the Constitution is belied by a review of other federal laws providing similar procedures, at least one of which has a 150 year pedigree. The 512(h) procedure is also similar to discovery in advance of federal litigation pursuant to Federal Rule of Procedure 27, which finds its origins in the Judiciary Act of 1789. Further, the subpoena power provided in subsection 512(h) does relate to cognizable Article III controversies, namely potential copyright infringement action as well as a dispute between the copyright owner and the ISP over access to the subscriber information.
The claim that subsection 512(h) violates the First Amendment does not withstand scrutiny. Subsection 512(h) does not proscribe spoken words or expressive or communicative conduct, nor is there a realistic danger that it will significantly compromise a recognized First Amendment protection. Section 512(h) merely requires a service provider to identify a person who appears to be engaging in copyright infringement, a necessary step before the copyright owner can initiate legal action. That action may range from an email or letter demanding that the alleged infringer cease and desist from the unlawful conduct to the filing of a lawsuit for copyright infringement. Section 512(h) does not offend the First Amendment any more than the filing of a lawsuit for copyright infringement. In fact, it is an essential tool for a copyright owner who intends to file such a lawsuit. Moreover, indeed, section 512 imposes sanctions on those who misuse the subpoena power, which serve to provide a safeguard.
Although not addressed in the Government's briefs in intervention, I think it is important to put into context the privacy claims that some now put forward. Users of peer-to-peer networks are, by definition, opening their computers up to the world. There may be an illusion of anonymity to that activity, but we have come to learn that such connections can also make available the user's social security number, credit card numbers, and other vital information. By contrast, the 512(h) subpoena process typically involves disclosure to the copyright owner of no more than the subscriber's name, email address, phone number, and perhaps street address. This hardly seems like an invasion of privacy.
The DMCA represents a carefully crafted and balanced bargain which utilizes the incentives created by pre-existing doctrines such as secondary liability as well as enlightened self-interest to encourage all stakeholders to work cooperatively to realize the potential of the Internet while respecting legal rights. Some are now selectively challenging key components of that bargain, particularly in the context of peer-to-peer technology. Taken together, the positions of Kazaa and Grokster, along with the arguments now made by Verizon, if they prevail, will leave copyright owners with little or no remedy against the most widespread phenomena of infringement in the history of this country. We know from past experience with Napster and current experience with Kazaa and Grokster that without a judicial remedy, this infringement will not stop, regardless of the availability of lawful alternatives. It is thus incumbent upon this Committee and this Congress to see to it that if the judiciary fails to enforce the DMCA and therefore fails to provide the protection to which copyrighted works are entitled, the legislature does.