Second Circuit Ruling in Viacom v. YouTube Is a Bummer for Google and the UGC Community
By Eric Goldman
Viacom International, Inc., v. YouTube, Inc., 10-3270-cv, 2012 WL 1130851 (2d Cir. April 5, 2012). The companion case is the Football Association Premier League Ltd. v. YouTube, 10-3342-cv
After five years in the courts, the Viacom v. YouTube litigation has finally produced an appellate opinion. In my opinion, the result is a loss for Google/YouTube and the UGC community generally. While the court largely agrees with many of YouTube’s contentions (and the ruling of the lower court), it nevertheless revives the litigation, ensuring that Google will spend millions of dollars more over the coming months/years.
Furthermore, the opinion identifies at least four “holes” in 512(c) coverage that future plaintiffs will surely attempt to exploit. (Smoking-gun internal emails; willful blindness; right and ability to control; and content syndication). This ensures that other UGC websites will spend a lot of money upfront to try to shut down those holes and spend even more money in litigation to demonstrate that it avoided those holes. So, on balance, I’m characterizing this opinion as a loss for the UGC community because this ruling increases the industry’s costs even if the substantive contours of 512 don’t change.
Given that the Second Circuit expressly disagrees with the Ninth Circuit’s UMG v. Shelter Capital ruling on the right and ability to control issue, I expect future 512 cases will be brought in the Second Circuit, not the Ninth Circuit. I’m not sure if this conflict is strong enough to get the case to the Supreme Court if Google chose to try.
Most importantly, this opinion exposes a structural deficiency of the 512(c) safe harbor. The statute’s simply too long and detailed, and if a defendant fails to satisfy each and every element, the safe harbor is lost completely. This is reminiscent of military strategy and information security: the defense has to work equally well across its entire border, while the adversary can concentrate its attack and only has to succeed on one point of attack to win. The same is true with a 512(c) defense. So, it doesn’t matter that YouTube won most of the points of contention; if any single point of contention fails, YouTube’s 512(c) defense fails. As I’ve insisted before, this provides a good lesson for drafters of safe harbors and immunities–to work effectively, the safe harbors/immunities must be pithy and categorical, or else they create too many potential points of failure.
Even though Viacom won this ruling, I still don’t understand how Viacom is making progress towards any strategic objective that matters to it. Viacom long ago conceded that it didn’t object to YouTube’s practices after 2008 (after it got access to Content ID). Indeed, Viacom gets upset with YouTube when it removes Viacom’s posts made for marketing purposes. And Viacom just expanded its licensing arrangements with YouTube. At this point, Viacom is very clear that it doesn’t want YouTube to go away, nor does it want any structural changes to YouTube’s current practices. So what the hell is this fight about? Viacom might still look at this lawsuit for its cash value, but it’s hard to be sympathetic towards that or see that as a big strategic objective. Viacom might be looking at this case to establish favorable precedent, but it picked a well-funded and determined defendant to make the point, and the Second Circuit ruling–though opening the door for copyright plaintiffs seeking to disqualify 512(c) defenses–doesn’t contain a big broad pronouncement that would constitute a strategic win for Viacom. The whole lawsuit is a big waste for all concerned, and the fact the parties can’t settle this case after 5 years of costly trench warfare continues to baffle me.
[UPDATE: This post is unintentionally a little contrarian; many folks see this ruling as a win for Google and the Internet. One possible explanation for this difference is our starting baseline.
If your starting point is Viacom''s summary judgment brief and some of the extreme and ridiculous arguments it makes, this opinion is a stunning and quite satisfying rebuke of that brief and those arguments. The opinion bristles with hostility towards most of Viacom's central arguments.
However, my baseline is the Shelter Capital case, which addressed many of the same topics and already established, as Ninth Circuit law, most of the conclusions that people are viewing as wins in this opinion. Compared to the Shelter Capital baseline, this opinion creates additional holes in 512's coverage. That's why I view this opinion as a bummer. In the end, this opinion narrows the grounds of 512 fights, and that's a good thing; but because it opens up new grounds for plaintiffs to exploit, this opinion pushes back--by many years--the day when the safe harbor preempts copyright owners from bringing meritless yet ruinous anti-UGC lawsuits. To me, that's a clear loss for the Internet, no matter how many points of contention YouTube actually won today.]
Analysis of the Court’s Discussion
512(c)’s Applicability to Direct and Secondary Infringement
One of the most important rulings of the court comes in an off-hand remark with minimal citation support on page 33: “The District Court correctly determined that a finding of safe harbor application necessarily protects a defendant from all affirmative claims for monetary relief.” This is one of the most significant questions in 512(c) jurisprudence: does 512(c) cover only direct infringement, or both direct and secondary infringement? Most courts have assumed the latter without ever saying so, but here the court (echoing, but curiously not citing, the Shelter Capital case) explicitly says 512(c) applies to all flavors of infringement. This makes the safe harbor potentially dispositive to the case–if YouTube gets it, Viacom loses any claim to money damages and, at best, can only get a meaningless limited injunction.
Knowledge of Infringing Activity
The opinion defines three types of service provider knowledge about infringement that might disqualify the service provider from 512(c):
1) actual knowledge of specific acts of infringement. The court calls this a subjective knowledge standard.
2) “red flags” knowledge of specific acts of infringement. The court calls this an objective knowledge standard. I’ve long argued that the “red flags” standard has evaporated in practice because, objectively, it’s impossible for anyone other than the content owner to look at a specific item of content and determine if it’s legitimately posted or not. In fact, even content owners can’t figure this out for themselves; Viacom itself repeatedly mistakenly identified which items of its own content were properly or illicitly posted to YouTube. If the content owner can’t make that determination, I’ll argue that objectively no one else could do so either.
3) willful blindness towards specific acts of infringement. Because of the nature of willful blindness, by definition it occurs in situations where the service provider otherwise doesn’t have subjective or objective knowledge of the infringing activity. Unfortunately, the court doesn’t say what service provider activities would demonstrate willful blindness, and many of us are scratching our heads wondering how willful blindness can occur when the service provider lacks actual or red flags knowledge. Combined with the Tiffany v. eBay contributory trademark ruling, it shows the Second Circuit is obsessed with willful blindness (though it didn’t define willful blindness there either—gee, thanks). The Ninth Circuit had a brief and oblique reference to willful blindness in the Shelter Capital opinion, but my guess is that plaintiffs will like the Second Circuit’s express discussion about willful blindness even better than they liked the Ninth Circuit’s casual reference.
The court identifies three pieces of evidence that YouTube may have had disqualifying knowledge:
* emails from Patrick Walker asking the team to look for and remove Football League clips
* an email from Jawed Karim indicating that there were Viacom clips on the site that may have been “blatantly illegal”
* an email exchange between Chad Hurley and Steve Chen debating whether to remove clips now or later.
The court expressly says that this evidence may not be enough for Viacom to show disqualifying knowledge (see FN9), but it is enough to get to a jury.
More generally, these emails remind us that YouTube was an unsophisticated start-up in its early days. They didn’t have legal counsel reviewing these emails or answering its questions about clip removals. Most UGC start-ups now know that these conversations shouldn’t be taking place over email, there should be a tightly enforced email retention policy, and active legal counsel is essential from day 1. But the modern protocol also mean that launching defensible new UGC start-ups is much more expensive.
This is especially true for UGC start-ups trying to avoid the willful blindness doctrine; I criticized the Tiffany v. eBay opinion for endorsing eBay’s very expensive anti-infringement infrastructure and implicitly requiring start-ups to maintain a similarly expensive infrastructure. This opinion may have the same adverse economic consequences for other UGC start-ups trying to minimize allegations of willful blindness in the copyright context.
Because the ruling creates more ways for plaintiffs to get to a jury in 512 cases, this ruling also means 512 litigation–even if the defendant succeeds–is going to be more expensive. The needle-in-haystack hunt for smoking gun emails means both parties will spend a lot on discovery (a point I complained about in the Shelter Capital case too). Furthermore, with respect to willful blindness, unless courts aggressively police plaintiffs’ allegations, it seems like plaintiffs can use a willful blindness allegation to defeat a 12(b)(6) dismissal motion; and if they can find any colorable evidence, plaintiffs can use that to defeat summary judgment and force jury trials in many future 512(c) cases.
[UPDATE: the willful blindness discussion in this case, and our confusion about what might possibly constitute it, reminds me a little of Justice Scalia's "tertium quid" reference in Walmart v. Samara. By adding another category of knowledge but not defining it, this opinion adds substantially to the transaction costs for a category that may not exist in the real world. The judges may be happy with themselves that they've done a more thorough taxonomizing job, but everyone else is miserable trying to figure out what belongs in that taxonomical node.]
Right and Ability to Control
The court blazes its own trail on what constitutes a service provider’s disqualifying “right and ability to control” infringing activity. It disagrees with YouTube, the lower court and the Ninth Circuit’s Shelter Capital case, all of which held that a service provider’s right and ability to control only applied when the service provider had specific knowledge of the infringing activity. But the Second Circuit also disagrees with Viacom’s proposition that “right and ability to control” imports its meaning from the common law vicarious infringement test. The court rightly recognizes that would render the statute internally contradictory.
So the court agrees with no one. Given that it rejected everyone else’s definitions, we might expect the court to carefully lay out what it thinks the phrase means. Sadly, no. The opinion doesn’t provide an express definition of what qualifies as the “right and ability to control,” instead sending that issue back to the district court to figure out both the standard and whether YouTube met it. The clearest clue the court provides about the standard is it “involve[s] a service provider exerting substantial influence on the activities of users, without necessarily—or even frequently—acquiring knowledge of specific infringing activity.” I have no idea what that means, other than that it’s open season for plaintiff fiestas.
[UPDATE: Sherwin Siy and I had an exchange on Twitter about this:
@SherwinPK: I don't think that "exerting substantial influence on activities of users" is "open season." A bit vague, yeah, but bounded
@ericgoldman OK, what you think "exerting substantial influence on activities of users" means? That's the goal of every UGC site!
@SherwinPK Not at all. Cybernet, for instance, had the OSP literally giving pointers on layout and content. That's substantial influence.
@SherwinPK Youtube doesn't exercise an editorial function wrt to user videos. That's why there's so much dreck.
My private reply: "That's not how the plaintiffs will put it! They will argue top X lists and exhortations to post constitute "substantial influence""]
Stored at a User’s Direction
The court rejects several of Viacom’s arguments that YouTube’s automated handling of user-supplied videos wasn’t stored at the user’s direction, including YouTube’s transcoding and playback functions and its display of thumbnails in a “related videos” module. However, the court leaves open the possibility that YouTube’s “syndication” of user videos didn’t qualify for 512(c). Specifically, YouTube licensed 2,000 user videos to Verizon Wireless. It’s unclear if any Viacom videos were included. That fact question goes to trial. If no Viacom videos were included, Viacom won’t get any benefit from this exception. However, the ruling leaves open future fights over what constitutes “syndication” as a way of bypassing 512(c). More plaintiff fiestas.
* Public Knowledge amicus brief in support of YouTube.
* Professor Michael Carrier’s amicus brief in support of YouTube.
* National Venture Capital Association amicus brief in support of YouTube.
* National Consumers League et al amicus brief in support of YouTube.
* NAMAC et al amicus brief in support of YouTube.
* MP3Tunes amicus brief in support of YouTube.
* IP and Internet Law Professors amicus brief in support of YouTube.
* Human Rights Watch et al amicus brief in support of YouTube.
* EFF et al amicus brief in support of YouTube.
* eBay et al amicus brief in support of YouTube.
* Consumer Electronics Association amicus brief in support of YouTube.
* CCIA/NetCoalition amicus brief in support of YouTube.
* Anaheim Ballet et al amicus brief in support of YouTube.
* BMI et al amicus brief in support of Viacom.
* Business Software Association amicus brief in support of Viacom.
* Microsoft/EA amicus brief in support of Viacom.
* Advance Publication et al amicus brief in support of Viacom.
* Brotman/Cass/Nimmer amicus brief in support of Viacom.
* Washington Legal Foundation amicus brief in support of Viacom.
* Seven IP professors’ amicus brief in support of Viacom.
* International Intellectual Property Institute amicus brief in support of Viacom.
* Eight professors’ amicus brief in support of Viacom.
* American Federation of Musicians et al amicus brief in support of Viacom.
* Vobile amicus brief in support of neither party.
* Audible Magic amicus brief in support of neither party.
* APILA amicus brief in support of neither party.