First, though, a few notes on what the opinion does not do:
- Judge Stanton's opinion is not definitive. It holds only between these parties, on these facts. It does not have much (if anything) to do with cases now proceeding against Veoh and other online-video sites. It does not "overrule" or even undermine Grokster and other anti-file-sharing cases. Most particularly, it does not act as a liability shield for the actual (alleged) infringers.
- This opinion has little or nothing to do with the fair use defense, 17 U.S.C. § 107. This is for both procedural and substantive reasons: Procedurally, the fair use defense had yet to be presented; and substantively, it's a losing argument for both sides that they desperately want to avoid getting into, because it is a fact-specific inquiry and there are too many alleged infringements.
- The opinion does not establish a wide-ranging defense for third parties against infringement claims that they did not already have... if they follow the rules. See Ellison v. America Online, Inc., 357 U.S. 1057 (9th Cir. 2004) (holding that AOL's administrative failures may have forfeited its right to the DMCA safe harbor) (n.b. I was lead counsel for the successful appellant).
This seems to beg the question a bit, though: What does Judge Stanton hold?
The background for Judge Stanton's ruling comes from the DMCA's safe-harbor provision, 17 U.S.C. § 512(c). (Unlike in Ellison, there was little controversy over whether these were "stored" or "transient" communications.) Judge Stanton framed the issue that he was actually deciding like this:
[A] jury could find that the defendants not only were generally aware of, but welcomed, copyright-infringing material being placed on their website. Such material was attractive to users, whose increased usage enhanced defendants’ income from advertisements displayed on certain pages of the website, with no discrimination between infringing and non-infringing content.
* * *
However, defendants designated an agent, and when they received specific notice that a particular item infringed a copyright, they swiftly removed it. It is uncontroverted that all the clips in suit are off the YouTube website, most having been removed in response to DMCA takedown notices.
Thus, the critical question is whether the statutory phrases “actual knowledge that the material or an activity using the material on the system or network is infringing,” and “facts or circumstances from which infringing activity is apparent” in § 512(c)(1)(A)(i) and (ii) mean a general awareness that there are infringements (here, claimed to be widespread and common), or rather mean actual or constructive knowledge of specific and identifiable infringements of individual items.
Slip op. at 6–7. (The failure to properly designate an agent was AOL's downfall in Ellison.) In a metaphysical sense, Judge Stanton has framed the inquiry as "What does it take to be 'aware' of infringing material on one's own system?"
And now, we're off to legislative-history land. Judge Stanton does make what is probably a harmless error... but, as this opinion is going to be cited repeatedly by people (including, unfortunately, judges) who have not read every word of the voluminous legislative history of the DMCA, I'm throwing up a reference here to point it out. Very simply, the error is this: Judge Stanton draws the wrong conclusion from the "red flag" language in the committee reports2 by failing to note the coordinate language concerning "sophisticated pirate directories" found slightly later in the House report, and by failing to reconcile that with the "holders need not monitor" counterpart of "ISPs need not monitor." In short, he makes a too-broad general statement that will no doubt be cited for particular propositions that are not what he is considering on this record.
That said, this case really comes down to distinguishing away Grokster and Napster, and in particular distinguishing between "significant infringing activity" and "rampant, dominating, intentional infringing activity." That's what Judge Stanton's opinion really does:
[] I make no findings of fact as between the parties, but I note that plaintiff Viacom’s General Counsel said in a 2006 e-mail that “… the difference between YouTube’s behavior and Grokster’s is staggering.” Defendants asserted in their brief supporting their motion and Viacom’s response does not controvert that:
It is not remotely the case that YouTube exists “solely to provide the site and facilities for copyright infringement.” … Even the plaintiffs do not (and could not) suggest as much. Indeed, they have repeatedly acknowledged the contrary.
The Grokster model does not comport with that of a service provider who furnishes a platform on which its users post and access all sorts of materials as they wish, while the provider is unaware of its content, but identifies an agent to receive complaints of infringement, and removes identified material when he learns it infringes. To such a provider, the DMCA gives a safe harbor, even if otherwise he would be held as a contributory infringer under the general law. In this case, it is uncontroverted that when YouTube was given the notices, it removed the material. It is thus protected “from liability for all monetary relief for direct, vicarious and contributory infringement” subject to the specific provisions of the DMCA.
Slip op. at 22–23 (citations omitted).
This is consistent with the "pawn shop" model that I've been pushing as the appropriate framework since the bill was first drafted. In the real world, there is probably no pawn shop in the nation that does not have a piece of stolen merchandise somewhere in it. Conversely, there are some pawn shops that are little more (if that) than fencing operations, knowing that virtually everything in the shop is stolen merchandise... even if the pawn shop owner carefully avoids ever learning from whom, or asking questions about particular pieces. On the facts of Grokster, Napster, Aimster, etc., those pawn shops were fencing operations. On the facts presented to Judge Stanton in this matter, YouTube is not, and a simple examination of virtually any search on YouTube demonstrates that that is a tenable conclusion... if only because so many of the clips on YouTube have arguable fair-use defenses to the particular clip. In the realm of "pawn shops," this is closer to determining whether the penurious son was really authorized to pawn his grandmother's silver candlesticks than to whether the known thief walking in with fine jewelry probably stole it in order to pawn it... and remembering that the question at issue concerns not the bailor's potential liability, but the pawn shop's.
Too, there's another reason that YouTube is different — for most purposes — from the music-infringers: Even at its highest quality, YouTube's material generally is not a complete, commercial substitute for the original... and those instances in which it is often seem to be those with the best fair-use claim. Consider, for example, the piece I've linked with this paragraph (using a custom link that doesn't pull in a lot of other BS). This is a relatively high-technical-quality audiovisual interpretation (sort of Ken Burnsish) of a relatively high-technical-quality audio recording. Whether it would have a winning fair use defense or not, it certainly has a tenable one... particularly given the result earlier this morning in the Germany v. England match at the World Cup <vbeg>. Judge Stanton's decision implies that when a high-enough proportion of material hosted by an ISP has a colorable fair-use defense, and that ISP has otherwise followed the DMCA's rules, the ISP is not a mere fencing operation, but an actual pawn shop operating a legitimate business in a dubious business environment...3 and that on these facts, YouTube was such a pawn shop, and therefore entitled to the DMCA's safe harbor.
The bottom line of Judge Stanton's opinion is that the pawn shop isn't liable for the sins of the son, however upset granny is about the candlesticks. That says nothing, however, about whether Inspector Lestrade might not haul said son off for "rigorous interrogation"... and it provides no protection at all for Jack the Fence's operation two streets over.
- This was a consolidated ruling that also applies to the parallel suit by the Premier League against YouTube. See Viacom Int'l, Inc. v. YouTube, Inc., No. 07civ2103 (S.D.N.Y.), and The Football Association Premier League Limited v. YouTube, Inc., No. 07civ3582 (S.D.N.Y.), consolidated opinion issued 23 Jun 2010.
- This is in part because he's not referring to the correct legislative history. He relies upon the House Commerce Committee report, H. Rep. No. 105–551 part 1 and (particularly) part 2. This report, however, concerned an earlier, radically different version of the bill. The Senate report (Sen. Rep. 105–190) and Conference Committee report (H. Rep. 105–796) concern versions of the bill that were much, much closer to what actually became § 512 (and the actual bill as passed in the Conference Committee report). Fortunately, Judge Stanton was careful to rely on material in Report 105–551 that had significant backup from the Senate report, but I am concerned that he did not note the virtual absence of any consideration of "red flags" in the Conference Committee report.
On this factual record, that does not appear to be a significant error. On most factual records, though — and Ellison was an example — it would be.
- We'll leave aside for the moment the rampant infringement of creator rights that occurs within the various industry areas that were/are suing YouTube in this instance. Viacom itself is not exactly a paragon of copyright virtue; and the less said about the Premier League's misuse of trademark law regarding fanstuff, the better. In short, while we're closer to questioning the right of the son to pawn granny's silver candlesticks than to dealing with an outright fencing operation, we still have to question his right to do so; we just don't arrest the pawn shop owner while doing so.