A while back, I posted my early reaction to the SDNY’s decision in the long-running Viacom lawsuit asserting secondary copyright infringement on the part of Youtube. Here’s what I said then:
The case was/is enormously important — Youtube was asserting that it was immune under the Digital Millennium Copyright Act (specifically, Sec. 512(c)) from copyright infringement claims arising out of user postings. Sec 512(c) sets up a “notice-and-takedown” scheme under which website owners are immune from third-party infringements as long as they “respond expeditiously” when notified of specific infringements by copyright holders. Viacom was relying on a portion of the statute that denies the immunity if the website operator has “actual knowledge that the material . . . on the system or network is infringing” OR if “in the absence of such actual knowledge, [it is] aware of facts or circumstances from which infringing activity is apparent . . .”
The critical question in the case was: given that it is a matter of common knowledge that there’s lots and lots and lots of infringing activity on Youtube, does that make the infringing activity “apparent”? If so, the 512(c) immunity is unavailable for Youtube. The court — correctly, in my view — said no, it does not. The “facts and circumstances” to which the statute refers must be of “specific and identifiable infringements of particular items. Mere knowledge of prevalence of such activity in general is not enough.” Sec. 512, and the other immunities provided in the Act for online conduct, “place the burden of policing copyright infringement — identifying the potentially infringing material and adequately documenting infringement — squarely on the owners of copyright. We decline to shift a substantial burden from the copyright owner to the provider . . .”
One could easily argue that these copyright immunities in the DMCA were a critical feature allowing “Web 2.0″ and “user-generated content” sites (like Youtube, Facebook, Myspace, and many, many others) to flourish. This opinion (though it will probably be appealed) goes a long way to protecting those sites from further attack by the copyright police. Nice work, Judge Stanton!
In the aftermath of the decision, there’s been, predictably, a backlash from commentators inside (and occasionally outside) the entertainment industry, arguing either that the decision’s not really that important, or that it’s just plain badly reasoned. Ben Sheffner’s piece over on Copyrights & Campaigns criticism of the opinion is thoughtful, but ultimately wrong-headed. Sheffner makes an interesting observation:
Section 512(c) of the DMCA identifies two triggers for the obligation of the host to remove the subject material (if it wants to maintain the safe harbor). First is actual knowledge of infringement (which can be obtained through receipt of a facially valid takedown notice pursuant to Section 512(c)(3)). Second is where the host becomes “aware of facts or circumstances from which infringing activity is apparent.” Id. § 512(c)(1)(A)(ii). This latter situation is known as “red flag” infringement; the idea is that the host can’t claim the safe harbor if red flags are being waved in its face, suggesting the obvious presence of infringing activity. The Ninth Circuit gutted the red flag doctrine in Perfect 10 v. CC Bill, specifically in this thoroughly unconvincing paragraph:
Perfect 10 alleges that CCBill and CWIE were aware of a number of “red flags” that signaled apparent infringement. Because CWIE and CCBill provided services to “illegal.net” and “stolencelebritypics.com,” Perfect 10 argues that they must have been aware of apparent infringing activity. We disagree. When a website traffics in pictures that are titillating by nature, describing photographs as “illegal” or “stolen” may be an attempt to increase their salacious appeal, rather than an admission that the photographs are actually illegal or stolen. We do not place the burden of determining whether photographs are actually illegal on a service provider.
In other words, under Ninth Circuit precedent…, having material identified by its poster as “illegal” and “stolen” is not a red flag that infringing activity is taking place. One is left to wonder whether the panel would have ruled the same way had actual red flags been waved in the defendants’ faces.
Judge Stanton, incorrectly in my view, adopted CCBill’s holding without much analysis, further rendering red flag infringement a dead letter. The statute (and legistlative history) clearly indicate that some form of knowledge beyond that imparted via DMCA notices qualifies as knowledge of “facts or circumstances from which infringing activity is apparent,” thus triggering a site’s takedown obligation (on pain of losing the safe harbor). But after reading Judge Stanton’s opinion several times, I simply have no idea what would actually constitute such “red flag” knowledge. And, again, his opinion does not even scratch the surface of the evidence presented by Viacom on this issue, see, e.g, Viacom Br. at 5-24, 50-56, and explain why none of it would raise a red flag for a reasonable service provider in YouTube’s position.
Sheffner’s absolutely right — the courts are indeed in the process of making the “red flag” exception disappear. And good riddance to it.
[Incidentally, there should be a name for this rhetorical phenomenon — where an opponent gives you, in the course of his criticism of a decision, ammunition for strengthening and even expanding its rationale. You see it a lot in dissenting opinions: “Under the majority’s reading of sections 543(c)(1)(ii)–(vi), all a defendant need show to escape liability is blahblahblah . . .” – and then defendants start to argue (even citing to the dissent) that because they can show blahblahblah, they shouldn’t be liable.]
Sec. 512 is a powerful defense precisely because it sets up a simple procedure: copyright owners find infringing material, they notify the host, the host takes it down, and the host is immune from liability. All steps along the way easily verifiable. “State of mind” inquiries — wasn’t it “apparent” from all the surrounding “facts and circumstances” that such-and-such was infringing? shouldn’t the host have known that? — are out of place here; among other things, they’re preposterously ill-suited to the scale of this problem. It would, I read recently, take you 15 years to watch the content uploaded to Youtube in a single day. What is, or is not, apparent to Youtube’s operators from this avalanche of material is not something we want to be arguing about, and it is not something I want web hosts to be worrying about. Judge Stanton got it spot on — the web host has no duty to act until the copyright owner identifies specific infringing material.
Do we want a powerful defense like this against claims of secondary copyright infringement on the Net? Damned right we do. I made the point in my earlier posting, and I think it’s interesting enough to reiterate. Virtually all of the wildly-successful 3d-party content sites on the Net — Youtube, Facebook, Twitter, Myspace, Flickr, Google News, Wikipedia, Blogger, . . . — come out of the U.S. Why is that? Why didn’t some college kid in the UK, or Italy, or Brazil, come up with the idea for Facebook? There are, I’m sure, lots and lots of factors at play — but I’m convinced that the existence of the immunity in sec. 512 in our copyright law is one of them. Even if Simon or Allessandra or Joao had the idea, and the technical wherewithal to pull it off, they’d be crushed before they got started – by many things, perhaps, but copyright liability is high on that list of crushers. Sec. 512 has meant that you can put your startup on line — even go to a bank or a VC and get financing — without worrying about potentially devastating copyright infringement liability. We don’t have many legislative successes in the copyright arena, so we should celebrate the ones we have.
[Thanks to Justin Gordon for one of the pointers]