Did a Court Eliminate 512(h) Subpoenas?–Maximized Living v. Google
By Eric Goldman with additional comments from David Gingras
17 USC 512(h) is a relic of a different era. The basic architecture of 17 USC 512 seeks to put copyright liability on users instead of their service providers. However, for that scheme to work, anonymous/pseudonymous infringers must be identifiable so the copyright owners can sue them instead of the intermediaries. 512(h) seeks to expedite the identification of alleged infringers by allowing copyright owners to get an unmasking subpoena super-easily. All copyright owners need to do is file a subpoena request with a court clerk, and in response the court clerk *must* issue the subpoena–the copyright owners don’t need to file a lawsuit, and no judge reviews or approves the subpoena’s issuance.
Indeed, neither the clerk nor a judge have any statutorily provided discretion to refuse the subpoena. As a result, 512(h) is now badly out-of-step with the law governing anonymous/pseudonymous online defendants that has developed over the past decade in response to unmasking abuses. In other areas than copyright, plaintiffs usually must make some showing that their substantive claims are meritorious before a judge will issue an unmasking subpoena. (The level of the plaintiff’s showing depends on a variety of factors). In contrast, a 512(h) subpoena issues irrespective of the substantive merits of the plaintiff’s claims–thus opening up a backdoor channel to unmasking abuses. For example, last year I got anecdotal reports that doctors used 512(h) to unmask patients that anonymously/pseudonymously reviewed doctors in contravention of the Medical Justice-supplied contract. If we were redrafting 17 USC 512 today, we would pay a lot more attention to 512(h) and its privacy implications than we did in 1998. [On that front, I have a latent empirical research project to investigate what happened after 512(h) subpoenas issued, but this case may have mooted it.]
With that background, let me turn to this case. Maximized Living sells copyrighted material to chiropractors. Anonymous blogger Doe allegedly infringed Maximized Living’s copyrights via a Blogspot blog post. Maximized Living submitted an apparently overbroad 512(h) subpoena request to Google to identify Doe, and Doe successfully quashed the subpoena for its irregularities. Nevertheless, Doe apparently removed the infringing material from the blog. After that removal, Maximized Living sent Google a putatively corrected 512(h) subpoena request to unmask Doe. In this ruling, the court quashes Maximized Living’s 512(h) subpoena for a second time.
The court does something goofy to reach this result. The court holds “that the subpoena power of s 512(h) is limited to currently infringing activity and does not reach former infringing activity that has ceased and thus can no longer be removed or disabled.” Thus, because Doe had removed the infringing material after the first 512(h) subpoena was quashed, there was no infringing activity taking place when the second 512(h) subpoena request was made.
The problem with this result is that copyright owners must submit a 512(c)(3) takedown notice to service providers before seeking a 512(h) subpoena. Most service providers will take down the allegedly infringing material in response to the 512(c)(3) notice, so unless the copyright owner moves really fast to make its 512(h) request, the infringing material invariably will be down before the 512(h) subpoena request gets filed with the court–leaving those copyright owner in the same place as this one (i.e., submitting a 512(h) request when there’s no current infringement). Below, David Gingras explains why the court may have misread the statute.
As a practical matter, this case’s result may not be earth-shattering even if it survives appeal. I believe most service providers honor 512(h) subpoenas without much scrutiny and perhaps without notifying the targeted individual. This case will only help if the targeted individual challenges the subpoena, which will only happen if the service provider notifies the individual before releasing the unmasking information and the individual gets to court quickly enough. Because the service providers are a critical player in this process, how they handle 512(h) subpoenas warrants careful attention. I’d be game to work with you to try to get service providers to tell us more about their 512(h) handling procedure and if they give notice to the users–and wait for any quashing effort to materialize–before forking over unmasking info. [FWIW, Google appears to have done both, so they get a gold star for the day.]
Copyright owners also can avoid this result by filing the 512(h) subpoena request basically at the same time as they send the 512(c)(3) notice. That way, when the 512(h) subpoena is filed, there is still infringing activity occurring, even if it’s quickly eliminated by the service provider responding to the 512(c)(3) notice. My guess is that many copyright owners will be reluctant to do this because it will increase the cost and time required to target infringing material when quick-filing of a 512(h) request will help in only a small number of situations. Thus, changing the takedown protocol to add a 512(h) filing probably isn’t cost-effective.
Finally, even if 512(h) isn’t available, the copyright owner can still seek unmasking through a John Doe lawsuit. This isn’t as low-cost as 512(h) and will trigger judicial screening of the subpoena request before issuance, so 512(h) is better for copyright owners if they qualify. Nevertheless, copyright owners can still achieve unmasking, and perhaps this case simply indicates that 512(h) is a much more highly specialized solution than we thought.
Finally, a personnel note: one of the plaintiff’s lawyers is Kenton Hutcherson. You may recall that last year I blasted an article by Kenton for advocating that plaintiffs scrub search results by taking advantage of Google’s apparently lax policy towards court orders. Here, it looks like the judge didn’t respond well to at least two of the plaintiff counsels’ choices:
1) the overreach in the initial 512(h) subpoena request
2) the submission of a second 512(h) without the court’s permission, as specified when the court quashed the first subpoena
One possibility is that the court reached its odd substantive conclusion in response to the plaintiff lawyers’ errors.
Comments by David Gingras
[Eric’s introduction: Many of you already know David Gingras due to his positions as General Counsel for Ripoff Report and litigation counsel for thedirty.com. While drafting this post, I sent this opinion to David for his thoughts, and his statutory analysis in response was so useful that I asked his permission to share it]
I think it’s extremely clear the court make the wrong decision here. I think the court should have found that the subpoena was entirely appropriate under § 512(h) even if the allegedly infringing material had been removed and the infringing activity stopped.
The court’s premise seemed to be that you could only use a pre-suit subpoena under § 512(h) to identify current infringers, not a former infringer who had stopped infringing. By itself, this seems like a very dubious distinction. What’s the difference?
As far as I can see, the conclusion was based on the fact that you obviously can only use what is commonly referred to as a “DMCA notice” (i.e., a takedown demand under § 512(c)(3)(A)) to address active infringements. In turn, that sounded correct because § 512(c)(3)(A) requires the party submitting the notice to identify, inter alia: “the material that is claimed to be infringing or to be the subject of infringing activity and that is to be removed or access to which is to be disabled.” By using the present and future tenses here, it’s beyond obvious that this section doesn’t apply to past acts of infringement. In other words, you can only use a § 512(c)(3)(A) notice to address current/ongoing infringements (DUH – if the material was already removed, you wouldn’t need to a send a takedown notice anyway, right?)
Up to this point, the court interprets the DMCA in a common sense way, but then it erred when it assumed (incorrectly), that because § 512(h) subpoenas are necessarily premised on a § 512(c)(3)(A) takedown notice, that requires the court to find that where the infringement has stopped, the right to pursue a § 512(h) subpoena also stops. That’s just totally inconsistent with the plain language of § 512(h)(5) which talks about the duties of a party on the receiving end of a DMCA notice (like Google) once they receive the follow-up subpoena:
(5) Actions of service provider receiving subpoena.–Upon receipt of the issued subpoena, either accompanying or subsequent to the receipt of a notification described in subsection (c)(3)(A), the service provider shall expeditiously disclose to the copyright owner or person authorized by the copyright owner the information required by the subpoena, notwithstanding any other provision of law and regardless of whether the service provider responds to the notification. [italics added]
The way I read that section, it seems pretty simple – you can get and serve a § 512(h) subpoena either contemporaneously with the § 512(c)(3)(A) takedown notice, or the subpoena may be issued subsequent to that notice; i.e., at a later time when the infringement has already stopped. Either way is perfectly fine, which makes sense.
In this instance, the way the court interpreted § 512(h) makes the words “or subsequent to” totally superfluous, so we know the court’s conclusion is incorrect. Furthermore, the last few words of § 512(h)(5) seem to suggest that § 512(h) subpoenas may or may not come after a service provider has already “responded” to the takedown demand; i.e., after the material has already been removed – that’s another strong indicator that the right to pursue a § 512(h) subpoena may start with a § 512(c)(3)(A) takedown notice, but it does not stop simply because the infringing material was removed.