The Digital Millennium Copyright Act (DMCA) insulates on-line service providers from copyright infringement claims if they:
This “safe harbor” from liability is the focus of two recent cases involving YouTube and isoHunt.com, a BitTorrent site. The business of YouTube is to enable the uploading and streaming of video content provided by users.
BitTorrent sites, in turn, help users identify files available for copying on third-party computers using peer-to-peer technology (think “Grokster” on steroids, but constructed to better distance the website from the infringing activity).
The ever-thorny problem is that a tremendous number of the videos and files that are available on or through these websites infringe the rights of copyright holders.
Both YouTube and isoHunt derive their revenue largely from advertisers, not from subscription fees. YouTube’s servers actually contain the infringing material; isoHunt merely directs users to peer-to-peer networks where it can be found.
YouTube’s internal studies estimate that 75% – 80% of all YouTube streams contain copyrighted material. IsoHunt has no such internal documents. Yet YouTube has so far escaped liability based upon the DMCA safe harbor, while isoHunt has not. What accounts for this different result?
These cases involve an active dialogue between two courts of appeals that cite each other in their opinions: the Second Circuit, which took up the YouTube appeal; and the Ninth Circuit, which heard the isoHunt appeal.
The two courts agreed on two key aspects of the DMCA safe harbor: The question of constructive knowledge of infringement (which they call “Red Flag Knowledge”) requires actual knowledge of facts and circumstances that would lead a reasonable person to conclude that infringing activity was taking place; and the element of control referred to in the third qualification for safe-harbor protection listed above must involve something more than the ability to locate infringing material and terminate users’ access to it.
In its lawsuit against YouTube, Viacom identified over 60,000 YouTube video clips that infringed its copyrights. In 2010, the trial court agreed to throw out the Viacom claims, holding that the DMCA safe harbor protected YouTube.
The Second Circuit Court of Appeals, in a closely watched 2012 decision, reversed the trial court and instructed it to apply the common law doctrine of “willful blindness” to determine whether YouTube was aware of the infringing activity. Thus, if YouTube was aware of a high probability of infringement and consciously avoided confirming it, YouTube might lose its DMCA immunity as a service provider.
In taking up the issue of willful blindness in its April 2013 ruling, the trial court noted that Viacom had advised YouTube of particular works that were being infringed, and had on occasion provided indications of where on YouTube to look for them.
However, Viacom did not identify the specific locations of infringements, leaving YouTube to find them. The DMCA specifically absolves service providers of any affirmative obligation to monitor the content posted by customers. Based upon that clause in the statute, the trial court was unwilling to tag YouTube with willful blindness.
By contrast, the owner of the isoHunt website actively urged his users to upload and download particular copyrighted works, provided assistance to those seeking to watch copyrighted films, and helped users burn copyrighted material onto DVDs. In its decision of March 21, 2013, the Ninth Circuit said that the content in question was sufficiently current and well-known that it would have been objectively obvious to a reasonable person that the material was copyrighted. For this reason, the isoHunt owner was deemed to have knowledge of the infringing activity.
Financial Benefit. IsoHunt receives its revenue from advertisers, not users. This fact might have been expected to protect IsoHunt from liability, given the statute’s requirement that the financial benefit be “directly attributable to the infringing activity.” However, the Ninth Circuit finessed this provision by pointing out that the owner of the website solicited advertising by pointing to the “TV and movies . . . at the top of the most frequently searched by our viewers,” and providing a list of typical user search queries, including popular movies and television shows.
The court also noted that a vast amount of infringing material was available through the use of the isoHunt websites – supporting an inference that the revenue stream is predicated on the broad availability of infringing materials for users, thereby attracting advertisers.
There is little discussion of the financial benefit issue in Viacom.
Control. In Viacom, the Second Circuit held that the ability to control infringing activity requires something more than the ability to remove or block access to materials posted on a service provider’s website. It struggled with how much “more” is required.
In the only previous case that found such control to exist, the service provider instituted a monitoring program by which user websites received “detailed instructions regard[ing] issues of layout, appearance, and content; forbade certain types of content and refused access to users who failed to comply with its instructions.”
After reviewing that precedent, the Second Circuit reversed the Viacom trial court because it had demanded proof of control over specific infringing activity, which the Court of Appeals found too restrictive for this prong of the test to determine eligibility for the DMCA shelter.
On remand, Viacom argued that YouTube exercised control over the infringing content by organizing and facilitating searches, by enforcing rules prohibiting pornographic content, and by removing some, but not all, infringing material. The trial court held that this was not enough control to deprive YouTube of the DMCA safe harbor.
The operator of isoHunt was not so lucky. The Ninth Circuit said:
. . . Fung organized torrent files on his sites using a program that matches file names and content with specific search terms describing material likely to be infringing, such as “screener” or “PPV.” And when users could not find certain material likely to be infringing on his sites, Fung personally assisted them in locating the files. . . .
. . . There is overwhelming evidence that Fung engaged in culpable, inducing activity. . .
Fung’s inducement actions. . . demonstrate the substantial influence Fung exerted over his users’ infringing activities, and thereby supply [the] component of . . . control.
On the surface, the Second and Ninth Circuits appear to be in harmony regarding the standards to be applied in measuring the availability of the DMCA safe harbor to on-line service providers. Each court agrees that failing to take down infringing content, despite awareness of facts that would cause a reasonable person to conclude that a specific act of infringement is occurring, can deprive the service provider of the safe harbor, as can deriving revenue from infringing activity over which the service provider exercises substantial control.
The Second Circuit went a bit further by stating that willful blindness can amount to knowledge. This test is slightly different from the “objective knowledge” standard described above because it requires evidence that the service provider made a deliberate effort to avoid guilty knowledge, and it incorporates a common law concept not found in the statute.
The trial court’s treatment of this question, dismissing the possibility of willful blindness because Viacom did not provide YouTube with the precise location of the infringed content, is unsatisfactory. If such notice were required for willful blindness, then what difference exists between actual knowledge and the imputation of that knowledge on the basis of willful blindness?
The trial court did not view the addition of search functionality as evidence of control, even though the Ninth Circuit felt quite differently about this question in considering whether isoHunt.com controlled its content.
Viacom has appealed the renewed summary judgment against it and can be expected to argue these aspects of the trial court’s decision. It will likely also point out the Ninth Circuit’s emphasis on the financial benefit accruing from the infringing activity, and its holding that advertising revenue can be directly attributed to infringement.
This case has attracted considerable attention, with content owners and their associations (BMI-ASCAP, for example) filing briefs supporting Viacom and public interest groups such as the Washington Legal Foundation filing briefs supporting YouTube. Given the money at stake (potentially multi-billions of dollars), the appeal will likely expose any weaknesses in the trial court’s summary judgment decision.
The key takeaway is that courts are willing to give lots of latitude to content-sharing websites, giving them a safe harbor from liability even when the majority of their content infringes copyrights. But when the content provider is purposefully directing its website and its users to the infringement of copyright-protected material, the malleable language of the DMCA will be construed so as to strip the service provider of the DMCA safe harbor.
Service providers whose activities have at least some legitimate, non-infringing purposes, and which are careful not to control the content that they make available, are likely to be safe from infringement claims, but those that are created purely to facilitate infringement are not.