Letter from AmeriKat II: Viacom v YouTube - The Viacom Appeal

Willfull Blindness cannot save you now

Viacom argues that, even if Section 512(c) excludes from liability those that do no not have URL-specific knowledge of infringement, the district court erred by finding for YouTube where the evidence showed that it was willfully blind to the massive scale of copyright infringement on the site. It was held In re Aimster Copyright Litigation (2003), cited in Arista Records v Doe 3 (2010) and recently in Tiffany v eBay that “willful blindness is knowledge in copyright law…as it is in the law generally.” Willful blindess occurs where a person engages in “deliberate avoidance” amounting to knowledge where “the circumstances were such to alert [the person] to a high probability” of the relevant fact, but the defendant “consciously avoided learning” that fact. A potential finding of willful blindness can be defeated where a defendant “continually taking steps to further refine its anti-fraud measures”, as was the case of eBay in Tiffany v eBay. Viacom argues that YouTube actually did the opposite of eBay in taking “affirmative steps to shut down any mechanism that might have provided the URL-specific knowledge YouTube claims is indispensable” for a finding of liability by removing the ability for community users to flag suspected infringing videos and only “selectively” implementing fingerprint technology. (picture, top left - YouTube's business policy? - "See no evil, hear no evil, speak no evil")

The Section 512(1)(B) Problem: YouTube Profiteering from Infringement

Section 512(c)(1)(B) requires that a defendant who benefits from a safe harbour protection must “not receive a financial benefit directly attributable to the infringing activity, in a case in which the service provider has the right and ability to control such activity.” Judge Stanton held that YouTube lacked the “right and ability to control” the activity because they did not have “item-specific” knowledge of the activity, but did not rule on whether or not they received a financial benefit. Viacom again argues that the touchstone of “item-specific” knowledge is not a pre-requisite for being able to control an activity. This is because, says Viacom, YouTube has the “ability” to control third-party infringement by implementing Audible Magic filtering.

Further, like with the district court’s interpretation holding Section 512(c)(1)(A)(ii) virtually meaningless, their interpretation of Section 512(c)(1)(B) is rendered likewise. This is because, if “right and ability to control” requires an ISP to have “item-specific knowledge”, this appears to be the same knowledge required for and ISP to fall foul of Section 512(c)(1)(A)(i) and (ii). Section 512(B), on this construction, essentially duplicates Section 512(c)(1)(A) as any service provider who has item specific knowledge of users’ acts of infringement automatically falls foul of Section 512(c)(1)(A) and thus never gets to Section 512(B).

Viacom argues that Congressional intent of Section 512(c)(1)(B) was to track the common law rule that a defendant may be found vicariously liable for copyright infringement where the defendant

“derive[s] a direct financial benefit from the infringement and ha[s] the right and ability to supervise the infringing activity.” (Ellison v Robertson (2004); Matthew Bender v W. Publishing Co. (1998))

This interpretation, says Viacom, is confirmed by Congress in the HR Rep NO 105-551(I) at 25-26, by the courts in Perfect10 and by academics in Nimmer on Copyright (section 12B.04[A][2] at 12B-38). Because vicarious liability turns on financial benefit and control, even in the absence of actual knowledge of infringement (Shapiro (1966); Grokster (2005)), YouTube’s activities clearly fall within the scope of Section 512(c)(1)(B). This is because, Viacom argues, YouTube has the right to control activities on the site by reserving editorial control in and the right to remove content and terminate accounts. It also has the ability to control the site by way of community flagging of suspected infringing videos, by way of its search feature and index and by implementation of fingerprint filtering technologies. YouTube also obtained a direct financial benefit attributable to the infringement because the infringing material acted as a “draw” or “major lure” for an ever-increasing YouTube audience. Such popularity resulted in YouTube being bought by Google for $1.65 billion only about 18 months after it was founded. This financial element is also seen by the placement of ads next to videos up until 2007.

No storage, no harbour!

A final requirement of Section 512(c)’s safe harbour is that the infringement is “by reason of the storage at the direction of a user” of the material. This includes service space for a user’s website, chatroom, or other forum where material is posted by users, which on the face of it may include a YouTube-type service. However, Viacom argues that its claims of infringement do not have anything to do with “storage” with or without the direction of a user: YouTube, in transcoding user-uploaded material into a standard format for display, distribution and performance of the content from its site, does not just facilitate storage but facilitates broadcasting. Viacom argues that a user’s decision to upload a video on to YouTube is not a direction to YouTube to then make copies of the video in different formats, to index and feature the material, or to licence the material to third parties to make viewing of the video easier on hand-held devices (such as the case with Verizon wireless). Viacom argues that YouTube takes those actions independently and for its own benefit and profit.

However, the AmeriKat cannot help but find fault with this argument because, if YouTube users just wanted their videos to be stored they would not be uploading them onto YouTube. YouTube users upload their videos on the site for the very reason that YouTube transforms the video into the particular viewable format. Surely their “direction” is implicit from the mere fact they are using YouTube to upload their video.

What’s Next

Almost 55 pages later, the bulk of Viacom’s substantive arguments end and its arguments for their motions for summary judgment on these issues begin—all in all 13,880 words or 15,000 more than this post.

A large swarm of amici curiae including Microsoft, The Washington Post, Newspaper Association of America, The Associated Press, and a group of economic professors also aligned with Viacom – a summary of their briefs can be found here. Microsoft’s brief in particular focused on YouTube’s “intentional efforts to build – and expand – its business based on a model that invited users to upload copyright infringing content to its site” and stating that this type of activity was not intended by Congress to benefit from the DMCA’s safe harbour provisions.

The AmeriKat is cautious about Viacom’s chances of success in their appeal and motion for summary judgment. Although she sees some logic in their arguments, especially in relation to redundancy of Section 512(c)(1)(A)(ii) with the heightened standard of knowledge, the US courts have in recent history been interpreting IP statutes in a way which benefits a Google, YouTube and eBay-type business model and she does not think that this case will be all that different. What do readers think? Will the Appeals Court save Viacom?

As far as the AmeriKat is aware, we have yet to receive a response from YouTube’s camp, be it press release or court filing. The AmeriKat has not been able to find any evidence of a peep or meow from YouTube, but if any reader knows of any, please let her know.

Letter from AmeriKat I: Viacom v YouTube - The Viacom Appeal

The past three Sundays the Amerikat has found herself engaged in many battles. First she was battling London Heathrow's snow incompetency, which delayed her flight home by four days. Then she was left battling a furiously stubborn cold, and then finally she was battling her over-stuffed suitcase in preparation for her trip to Singapore for the Global Foum on IP (more about this later this week). Having flown around the world in four days, she is now battling a bizarre jetlag cocktail. (picture, left - the AmeriKat with more legroom than she has experienced in weeks) However, the AmeriKat has pulled herself together this week with a report on a different battle: this time Viacom's opening brief filing in the Viacom v YouTube litigation, filed last month. For previous IPKat reports on the battle click here.


Viacom v YouTube: Viacom's appeal arguments

Following YouTube's summary judgment success in the Viacom litigation last year it was no surprise that Viacom, supported by several amici curiae, would appeal Judge Stanton’s June summary judgment (here and reported here). Sure enough, on 3 December 2010, Viacom filed its opening brief with the US Court of Appeals for the Second Circuit and requested summary judgment on the issues. Readers will recall that this case focuses on Viacom’s copyright-protected content being uploaded onto YouTube by third-party users and YouTube, once notified, not doing enough to remove the infringing content or in preventing the uploading of infringing material.

Judge Stanton held that YouTube was able to benefit from the Safe Harbor provisions in the Digital Millennium Copyright Act because it did not have actual or constructive knowledge of every specific YouTube URL containing infringing material on the site. Mere knowledge that there is, or is likely to be, infringing material on the site was not enough to be considered actual or constructive knowledge under Sections 512(c)(1)(A)(i) and (ii) respectively. To hold so would, Judge Stanton held, fly in the face of the DMCA's text, specifically Section 512(m)(1), as well as the Perfect 10 case—both of which indicate that the burden of identifying infringement must rest with the copyright owner, not on YouTube.

The Contentious Section

To refresh your memories, the section at issue is Section 512(c)(1) which provides that a service provider:

“shall not be liable for monetary relief or other equitable relief, for infringement of copyright by reason of the storage at the direction of a user of material that resides on a system if the provider:

(A) (i) – does not have actual knowledge that the material or activity using the material on the system or network is infringing; or

(ii) – in the absence of such knowledge is not aware of facts or circumstances from which infringing activity is apparent (the "red flag test"); or

(iii) – upon obtaining knowledge or awareness expeditiously removes it; AND

(B) – does not receive a financial benefit directly attributable to the infringing activity; AND

(C) – upon notification of claimed infringement it responds expeditiously to remove the material”

Viacom Sets the Scene

In its opening brief, Viacom calls Judge Stanton’s interpretation of Section 512(c) "absurd, disquieting and disruptive" and states that the effect of this interpretation would be that even the most

"piratical businesses held to account in Metro-Goldwyn-Mayer Studios v Grokster (2005) could be immune with just minor tweaks to their business models.”

Viacom also argue that the text of DMCA does the opposite to what Judge Stanton held, in that the DMCA actually "compels" internet service providers who are aware of pervasive copyright infringement, participate and profit from it, to be found liable for copyright infringement under the DMCA.

Viacom's opening brief first sets out the legislative history of the DMCA. The US Congress (picture, left) when drafting the DMCA, sought to achieve a balance between protecting copyright owners from mass online infringement and providing security to internet service providers who served as the "backbone" of the internet. In drafting Section 512, Congress aimed to strike this balance by creating a safe harbor for "innocent service providers" which disappears the moment that they lose their innocence (ALS Scan v RemarQ Communities (2001)). This disappearance of the safe harbor occurs when the ISP obtains actual or constructive knowledge (the "red flag test") of infringing activities and then does not act expeditiously to remove or disable access to the material. Congress also wove into the DMCA safe harbour, principles of vicarious infringement in order to exclude its shield in instances where ISPs receive financial benefit attributable to the infringement.

The brief legislative history segues into a somewhat damning historical account of YouTube's history entitled “YouTube Builds A Business Based on Infringement” which cites gems such as an email from YouTube founder Steve Chen who stated that removing the “obviously copyright infringing stuff” would reduce YouTube views “from 100,000 views a day down to about 20,000 views or maybe even lower”. From further internal YouTube emails, Viacom lifts other extracts which indicate that YouTube was aware of mass copyright infringement but chose to take a passive role in the self-monitoring of the site in order to benefit from the safe harbour legislation. Viacom also points to YouTube’s failure to implement software or digital fingerprint technology that would alert or prevent the upload of copyright material as being indicative of not only YouTube’s knowledge of the pervasiveness of infringement on the site, but as a conscious plan to evade liability by placing the burden of policing infringement squarely on the shoulders of a copyright owner in an effort to maintain their site traffic. In 2006, a “little exercise” conducted by Chen showed that over 70% of the “most viewed/most discussed/top favorites/top rated” videos on YouTube were copyright material. Viacom states that this method of business only grew after YouTube was bought by Google for $1.65 billion in late 2006.

The Section 512(1)(c)(A) issues: Surely YouTube knew what was going on!

Viacom contends that the district court erred in holding that YouTube could benefit from the safe harbour provisions, despite being “generally aware of” and indeed “welcome[ing]” of copyright-infringing material, because they lacked knowledge of the specific URL of each individual infringing video. Viacom argues that there is actual evidence that YouTube did have item-and-location specific information in respect of some of the works complained of and, for those that it didn’t, YouTube chose to actively to blind itself from assimilating specific information of infringement in order to benefit from the safe harbour, i.e. by not implementing infringement detecting software or a community flagging system for infringing videos. The internal email evidence also indicates that, for years, YouTube’s policy and practice was to take “no action”. Viacom argues that because YouTube, in the district court’s own words, “welcomed” “blatant” infringement, and turned a blind eye to such infringement it knew was occurring as evidenced by the internal emails, it cannot be said to not be “aware of facts or circumstances from which infringing activity is apparent” under Section 512(c)(1)(A)(ii). A defendant’s inaction while aware of widespread incidents of infringement, Viacom argues, should not be rewarded with a shield from liability.

That being said, the district court said that the facts above were not specific enough for YouTube not to benefit from the safe harbour provision. However, Viacom argue that the district court’s higher standard of knowledge for Section 512(c)(1)(A)(ii) is not supported by the statutory language. Breaking down the provision Viacom states that the exclusion of the safe harbour depends upon the defendant’s awareness of “facts or circumstances” that make the “infringing activity” “apparent”. This does not mean that these facts have to automatically point to an activity being absolutely and conclusively illegal (citing Jane Ginsburg’s 2008 article in 50 Ariz. L. Rev. 577). Further “facts and circumstances” suggests that Congress intended there to be a more holistic view of the origin, quality and quantity of information of infringement which a defendant may possess to fall within or foul of the provision. That is to say there is not just one type of specific information that is required for the inoperability of the safe harbor under Section 512(c)(1)(A)(ii) ( i.e., the requirement of specific URL addresses) but a combination of information sufficient to raise a “red flag” of warning to the service provider. Holding that there has to be specific identifiable knowledge on the part of the defendant to find an ISP liable renders the purpose of 512(c)(1)(A)(ii) void. If the same standard of knowledge is required for 512(c)(1)(A)(i) as for (ii), what purpose does (ii) serve?

Turning away form statutory interpretation, Viacom argues that the district court’s interpretation cannot be reconciled with Congress’s intent in enacting DMCA. A main purpose of the DMCA was to provide reasonable assurance to copyright owners that their copyright would be protected, given the increased risk and ease that their works can be infringed online. Although Viacom has a hint of recognition that the safe harbour provisions in the DMCA are appropriate for “innocent service" it argues that if the DMCA is to

“conform to the central purposes of the statute, [it] must exclude at least those that ‘welcome’, and even intend, their users’ infringement. To conclude otherwise would fatally undermine Congress’s intent to address 'massive piracy.' It would immunize even entities such as Grokster itself, which ‘distribute[d] a device with the object of promotion its use to infringe copyright’, yet designed its system to avoid item-or location-specific knowledge of those infringements.”

Viacom argues that YouTube has not proved that Congress intended the DMCA to be anything other than this original intent and indeed cites another Second Circuit district court decision as recognizing the DMCA’s purpose as being just that (Arista Record LLC v USENET.com, Inc (2009)). To hold otherwise would also place a substantial burden on copyright owners, who would have to continuously and constantly monitor the entire site for infringing videos.

Judge Stanton’s decision also relied on the district court’s interpretation of Perfect 10 v CCBill (2007) where the defendants provided services to websites with domain addresses such as “illegal.net” and “stolencelebritypics.com”. The claimant argued that these domain names gave notice to the defendants of ongoing infringement on these cites. The Ninth Circuit held that the website names alone were insufficient to create awareness of infringement as the names may just be a method of increasing traffic to the site, rather than being conclusively illegal. Viacom argues that

“even the crabbed (picture, above left) construction of red flag awareness in the Ninth Circuit’s analysis would not save YouTube, however, for, the service provide in CCBill was found to have no awareness that infringement was ongoing at all. Here, in contrast, YouTube was well aware that massive infringement as occurring, intended it to occur, and made no attempt to remedy it."

Procedural irregularities: when do you let go of a decision?

The IPKat's friend, IP blogger, commentator, practitioner and enthusiast Barbara Cookson (Filemot Technology Law Ltd) writes to draw his to attention to an unusual episode in the annals of the British Trade Mark Registry. Barbara writes as follows:

"This was a fairly standard opposition decided on the papers, with the primary ground based on similarity with an existing registration.

That existing registration was over five years old, so proof of use was required and duly filed. Since Hackett is a high-street brand, they were able to show use of their H logo on aftershaves, cufflinks, silverware, leather goods and clothing. This meant there was an overlap of identical goods between the two marks. On 17 July 2009, in MIP Metro Group Intellectual Property GmbH's international registration; opposition of Hackett Ltd, Case O-210-09, Ann Corbett applied a global approach and came to the conclusion "albeit with some hesitation" that there was no likelihood of either direct or indirect confusion, even where identical goods are concerned (paragraph 51).

There were also grounds under section 5 (3) and (4) (passing off and dilution), both of which require evidence of reputation/goodwill. Ms Corbett's view of the evidence was that it did not allow her to find that the mark had a reputation and both these grounds were swiftly dismissed primarily because the marks were not similar in her view. The result: victory for MIP and a costs order of £1700.

In MIP Metro Group Intellectual Property GmbH's international registration; opposition of Hackett Ltd, Case O-348-09, on 4 November 2009, Oliver Morris issued a second substantive decision, stating in paragraph 6 that this was because the earlier one had been set aside by the registrar due to procedural irregularity (not set out in the decision, but believed to have been the fact that Ms Corbett did not have all the submissions from the parties). The evidence mentioned by Mr Morris is the same as that mentioned by Ms Corbett. Mr Morris arrived at a slightly more limited specification for the earlier right. The silverware went out. The leather goods became specific and the clothing was only men's. At the end of his global assessment -- which is pretty lengthy and deeply analytical, Mr Morris did find that there was a likelihood of confusion for the goods which were identical or highly similar (aftershave and clothing).

Mr Morris also gave rather more attention to the passing off and dilution grounds than Ms Corbett, but eventually came to the same conclusion that they failed because the evidence was not strong enough. Since Mr Morris had found for Hackett on a few goods he decided not to make a costs order.

Both decisions seem to be well within the normal range of conclusions that a hearing officer could draw on this scenario".

The IPKat is fascinated. In particular, though he is unfamiliar with the highways and byways of registry practice in instances such as this, he wonders whether readers are happy with signed and published decisions being set aside in circumstances other than the good old-fashioned appeals he enjoys so much. He believes that this case has indeed gone to appeal-- which is where many people might consider it should have gone in the first place.

There's another problem with in-office procedural regularities. We don't know about them unless the office confess to them. I trust your skills to create some interest. At the very least we will open the floodgates and any agent who has an opposition decision they don't like will protest. What fun! Merpel adds (not that it makes any difference) that MIP Metro uses an H on its H-line hotel products, which it sells through Makro stores.