Secondary Liability for Copyright Infringement in the US

Secondary Liability for Copyright Infringement in the U.S.: Anticipating the Aprés-Grokster
by Professor Jane Ginsburg

illustration: David Arroyo

On the last day of the 2004 Term, the Supreme Court announced its much-awaited decision in MGM Studios, Inc. v. Grokster Ltd., 125 S. Ct. 2764 (2005). Songwriters, record producers and motion picture producers alleged that two popular file-"sharing" networks, Grokster and Streamcast (dba Morpheus) should be held liable for facilitating the commission of massive amounts of copyright infringement by the end-users who employed the defendants' peer-to-peer (P2P) software to copy and redistribute films and sound recordings to each others' hard drives. The Court reversed the Ninth Circuit's grant of summary judgment for defendants, holding that the technology entrepreneurs could be held liable for "actively inducing" the end-users' acts of infringement. As consumer-wielded digital media increasingly supplant the traditional intermediaries who made copyrighted works available to the public (and who traditionally were the targets of copyright enforcement), courts have struggled to balance meaningful protection for works of authorship against the progress of technological innovation. Grokster is the latest in a series of decisions to address that balance by articulating the liability of an enterprise that does not itself commit copyright infringement, but instead makes it possible for others to infringe. To appreciate the Supreme Court's analysis, it helps to set the case in doctrinal context.

Copyright infringement is a tort. So is enabling or inciting another to infringe, at least when the enabler knows that her conduct will result in infringement. Decisions dating back several decades recognize that one who supplies the means to infringe, and knows of the use to which the means will be put (or turns a blind eye), can be held liable for contributory infringement. In the early cases, however, the relationship between the supplier and the user of the means was sufficiently close, that there could be little doubt of either the knowledge or the nexus between the means and the infringement. For example, in the "make-a-tape" case, Elektra Records Co. v. Gem Electronic Distributors, Inc., 360 F. Supp. 821 (E.D.N.Y. 1973), a record shop rented sound recordings to customers who would also purchase blank tape and then use a recording machine on the store premises to copy the rented recording onto the blank tape. The store owner's knowledge of the likely use of the blank tape was patent. When, however, the infringement-facilitating device leaves the direct control of the facilitator, so that he no longer knows in fact what his customers are up to, contributory infringement may be more difficult to establish. That, in essence, was the copyright owners' problem in Sony Corp. of America v. Universal City Studios, 464 US 417 (1984) (also known as the "Betamax" case). Sony, the distributor of the videotape recorder, could well anticipate that consumers would use the record function to copy protected programs, but once the device was out of the manufacturer's hands, it could neither know precisely what the end users were doing, nor limit their use to permissible copying.

In absolving Sony of liability, the US Supreme Court added a gloss to the prior standard: one who distributes an infringement-enabling device will not be liable for the ensuing infringements if the device is "widely used for non-infringing purposes. Indeed it need merely be capable of substantial non-infringing use." This is so even though the distributor was aware that at least some of the use to which the device would be put would be infringing. The Court then held that time-shifting (recording for subsequent viewing and then erasure) of free broadcast television programs was a fair use. On the record in the case, the "primary use" of the VTR was for time-shifting. A use held to be non-infringing thus predominated, and certainly met the "substantial" standard.

The Sony "substantial non-infringing use" standard again came into play with respect to mass-market means of copying in A&M Records, Inc. v. Napster, Inc. 239 F.3d 1004 (9th Cir. 2001). There, an online peer-to-peer music "sharing" service maintained a central database that allowed end users to find other users currently online, and to copy MP3 files from their hard drives. Napster evoked the Sony standard, asserting that not all the files were copied without authorization. Napster also asserted that P2P architecture could in the future spawn more non infringing uses. The Ninth Circuit agreed that Sony required taking into account the service's capacity for future lawful use, but nonetheless held Napster a contributory infringer. In yet another gloss on the standard of liability, the Napster court held that courts should inquire into non infringing uses when the distributor of the device lacks actual knowledge of and control over specific infringements. Where, however, it is possible to segregate and prevent infringing uses, it is not appropriate to exculpate the entire system by virtue of its capacity for non infringing uses. In other words, the consequences to technology of enforcing copyright rules were different in Sony and in Napster. Sony presented the court with an all-or-nothing challenge: either the device would be enjoined, frustrating legitimate uses, or no liability would attach, despite the infringements the device enabled. In Napster, by contrast, the service could disable infringing uses by blocking access to listings of protected files, while allowing permissible uses to continue. Napster thus transformed Sony into an inquiry into knowledge of and ability to prevent specific infringements.

Of course, Napster set out the instructions for its own demise: if Napster was liable because it could maintain control over its users' activities, then the next device or service would make sure to make it difficult if not impossible for the service to exercise control. So were born Kazaa and its US licensees, Grokster and Morpheus. Unlike Napster, these services had no centralized directory: they dispersed information about file locations across computer "nodes" around the world. Users could find each other, but the services disclaimed the ability to prevent infringements as they were occurring. Although it recognized that Grokster and Morpheus had intentionally built their systems to defeat copyright enforcement, the Ninth Circuit held that, without the ability to prevent specific infringements, the services could not be liable. The court scarcely considered whether the services enabled substantial non-infringing use; it acknowledged that 90 percent of the uses were infringing, but observed in a footnote that 10 percent could be substantial, particularly when the 10 percent referenced many millions of uses. (That the other 90 percent would be even more extensive seems not to have troubled the court.)

A unanimous US Supreme Court reversed. It held that the Ninth Circuit had misapplied the Sony standard, or, more accurately, that the Ninth Circuit did not appreciate that the Sony standard does not even come into play when the defendant is "actively inducing" copyright infringement. That is, a device might well be capable of substantial non-infringing uses. But if it can be shown that the distributor intended users to employ the device in order to infringe copyright, then the distributor will be liable as a matter of basic tort principles. In this light, Sony was a case articulating a standard for assessing liability when it cannot be shown that the device distributor sought to foster infringement. But if the defendant has actively induced infringement, there is no need to revisit the Sony standard in order to clarify what "substantial non infringing use" actually means.

The Court set out three elements probative of intent to induce infringement: (1) the defendant promoted the infringement-enabling virtues of its device; (2) the defendant failed to filter out infringing uses; (3) defendant's business plan depended on a high volume of infringement. In Grokster's case, all three elements were easily demonstrated. Grokster had sent out emails extolling P2P copying, and it had "aim[ed] to satisfy a known source of demand for copyright infringement, the market comprising former Napster users." Grokster not only declined to devise its own filters; it blocked third-party filters. And Grokster's business plan depended on advertising whose rates would turn on the volume of users encountering the ads. The more Grokster could attract visitors, the better for business, and the prospect of free music attracts more visitors than paid music. Taken together, these factors demonstrated a clear intention to foster infringement. As the Court declared: "The unlawful objective is unmistakable."

Of course, inducement to infringe is actionable only if infringements in fact occur. Because the liability derives from primary infringing conduct, bad intent must join with unlawful end user acts. Thus, for example, distributing a copying device together with an exhortation to use the device to engage in massive unauthorized copying does not give rise to liability if no one in fact does so. In Grokster, however, end user infringement was never in doubt; plaintiffs' studies showed that 90% of the works copied were copyrighted, and even the Ninth Circuit acknowledged that the "Copyright Owners assert, without serious contest by the Software Distributors, that the vast majority of the files are exchanged illegally in violation of copyright law." The Court thus could exclaim: "The probable scope of copyright infringement is staggering."

Having ruled that bad intent, if proved, sufficed to attract liability for infringements thus induced, the full Court declined to analyze what the standard for contributory infringement would be when intent to foster infringement cannot be shown. The full court opinion provided some indication of that standard, however, when it stressed that certain of the three indicia of intent could not, in isolation, establish inducement, because basing liability solely on a defendant's business plan, or solely on the design of its product, would be inconsistent with Sony. But the Court assiduously declined to offer further guidance on the meaning of "substantial non infringing use."

But perhaps the failure to clarify Sony will not in fact matter very much, because the Grokster inducement standard may displace inquiries into the substantiality of non-infringing uses. Speculation is hazardous, but one might predict that, where a device or service facilitates infringement on a massive scale, its distributor will likely be found to have intended that result. Where the infringement the device or service enables is relatively modest in scale, inducement will not be found, but neither will the Sony threshold for liability be held to have been crossed, whatever its height. In other words, "inducement" and "substantial non-infringing use" will become legal conclusions, separating the Sony (good technology) sheep from the Grokster (evil entrepreneur) goats.

The following analysis supports the speculation. Suppose in each of the ensuing cases, the device or service made available to the public enables a very high volume of infringement.

Case 1. The distributor,, promotes the infringement-facilitating features of its device or service and does not filter out infringing uses. But it has no business plan because it does not seek revenues; rather it wishes to liberate content from copyright's clutches.
Although one of the three Grokster elements is missing, this still seems like a straightforward case of "inducement" and, assuming the device is used as intended, liability would be found.

Case 2. The distributor,, neither promotes infringement nor filters infringements out; its business plan requires a high volume of traffic to the site. Bloggers on the unaffiliated site identify and promote the infringment-facilitating features of's device or service.

The most probative Grokster element, advertising, seems absent here. If merely benefits from, but is not in league with,, then the question becomes whether failure to filter, plus an infringement-dependent business plan, suffice to establish inducement. Grokster tells us that either of these two in isolation will not, but Grokster did not explicitly require all three elements to be present before inducement could be found. Moreover, as the Australian Federal Court recognized in a lawsuit brought against the KaZaa P2P network, Universal Music Australia Pty Ltd v. Sharman License Holdings Ltd. [2005] FCA 1242, the distributor of a device or service is not likely to filter if to do so would reduce its economic benefit. In other words, the two go hand in hand. Other Grokster elements prove interdependent as well: a site that does filter is not likely to advertise an ability to facilitate infringement if it has in fact hampered that capacity. A site that does filter, moreover, will probably not be engendering massive infringement.

Grokster may instruct that technology entrepreneurs should not draw their start-up capital from other people's copyrights. Pressing copyright owners into service as the principal (uncompensated) investors in a new technology seems rather problematic. But what if there are other investors? That is, let's assume that lawful uses account for some portion of the device or service's appeal, so that the service could start up thanks to the revenues derived from the lawful uses, although input from other people's copyrights may remain necessary to profitability. The Grokster goats are those technologists whose business plans would not exist but for the infringements they enable. What barnyard characterization fits those whose business plans, albeit copyright-parasitic, contemplate mixing infringing and non infringing sources of revenue? If we take Grokster's treatment of the role of intent literally, then the partial foundation of the business plan (and the concommitant device design) on non-infringing uses should not matter; the entrepreneur still intends to foster some infringement. Liability, however, may turn on how much infringement. For we have posited that Grokster will supply the rule when the actual or potential volume of infringement is "staggering." Short of that - but how much short of that remains uncertain - the Sony standard reappears, with (as we have posited) the result that the challenged technology will be deemed lawful.

This essay was excerpted from an article by Professors Jane C. Ginsburg and Sam Ricketson (University of Melbourne Law School) titled "Inducers and Authorisers: A Comparison of the U.S. Supreme Court's Grokster Decision and the Australian Federal Court's KaZaa Ruling," in the Media and Arts Law Review (University of Melbourne, March 2006).