Photobucket Qualifies for the 512(c) Safe Harbor (Again)–Wolk v. Kodak

By Eric Goldman

Wolk v. Kodak Imaging Network, Inc., 2012 WL 11270 (S.D.N.Y. Jan. 3, 2012). Prior blog post on this case.

As I’ve indicated before, blogging 17 USC 512 cases has gotten tedious because they are just TOO LONG. I can crank through most 47 USC 230 cases in an hour or two because they are usually quite short and efficient. In contrast, because 17 USC 512 gives copyright plaintiffs so many words to contest, 512 opinions tend to be lengthy and quite time-consuming to blog–with this 69 page opinion as a prime example. This has some implications for drafters of laws like SOPA/PIPA, which have similarly long and detailed provisions that just beg plaintiffs to contest every word and will force courts to write quite lengthy opinions that bloggers like me will struggle to crank through. I cheer for immunities and safe harbors, but I have three cheers for SHORT immunities and safe harbors.

This case is even more unfortunate because the pro se plaintiff had an obviously unmeritorious case, yet the two defendants used three law firms to beat this case. And it’s not exactly like Kodak is wallowing in cash any more.

Wolk is an artist. Users uploaded images of Wolk’s work to Photobucket (a UGC photo-sharing site). Photobucket, in turn, had a revenue-sharing agreement with Kodak Imaging that allowed users to print the images via Kodak (i.e., Kodak did “photofinishing”).

Photofinishing Liability

The court says that Kodak Imaging wasn’t directly liable for printing the images (Wolk didn’t allege secondary infringement). The court observes that “reproduction, display or transmission of the Plaintiff’s images by or through the KODAK Gallery website is an automated process with no human intervention by any employee of the Kodak Defendants.” Thus, because its entire system was automated, Kodak didn’t act volitionally and thus avoids the strict liability standards of direct copyright infringement.

This ruling is unexpected because it’s been conventional wisdom for many years that photofinishers were in fact directly liable for their print jobs. Perhaps that’s because humans were always involved in the photofinishing process during that time, as opposed to now where the process from photo upload to mailing of items can be completely automated. Whatever the case, this ruling has to be encouraging for other automated photofinishers (whether they print photos or other items), such as CafePress or Zazzle. Then again, perhaps the copyright plaintiffs will pursue them under secondary infringement doctrines, which Wolk didn’t do.

Although I like the result, I remain confused about the scope of the “volitional doctrine.” As was the case in Cablevision, Kodak’s system was completely automated only because Kodak’s engineers designed it that way. We would benefit greatly from a richer theoretical grounding for the volitional doctrine and how it interplays with strict liability. Without that grounding, the results seem a little random.

512(c) Safe Harbor

Photobucket qualified for the 512(c) safe harbor. This isn’t surprising; the court indicated as much when it denied Wolk’s request for a preliminary injunction. Still, the court works through a 512(c) in fine detail:

* Photobucket is a “service provider”

* Photobucket properly adopted and implemented a repeat infringer termination policy.

* Photobucket accommodates standard technical measures. Wolk argued that Photobucket gives users tools that can remove or hide watermarks. The court doesn’t opine whether watermarks are a standard technical measure, but instead the court says Photobucket doesn’t encourage users to use the tools, so users–not Photobucket–would be the ones interfering with standard technical measures if watermarks qualified as such.

* Photobucket didn’t have actual or constructive knowledge of the infringement. Before the lawsuit, Wolk sent 15 infringement notices covering 9 works. When Wolk sent 512(c)(3) notices, Photobucket expeditiously responded. However, 11 of the notices weren’t 512(c)(3)-compliant (because they didn’t specify URLs) and thus are irrelevant. (Compare the troubling dicta in the uncited UMG v. Shelter Capital). Wolk argued–as so many copyright owners do–that one notice about a work should cover all existing and future uploads without providing URLs of the other items. The court rejects that argument.

* Photobucket doesn’t have the right/ability to control infringement because it does not prescreen content, render extensive advice to users regarding content and edit user content. Photobucket also lacked direct financial benefit from the infringement: “The Defendants’ profits are derived from the service they provide, not a particular infringement.”

* Photobucket properly identified its agent for notice and designated it with the copyright office.

All of this 512(c) analysis was fairly by-the-book. The most interesting part is where the court discussed how “Photobucket Has No Duty To Police Its Website For Infringements.” The court says:

Photobucket is a website that consists of over 9 billion images and videos. Under the plaintiff’s theory, Photobucket would be required to police its website for infringing copies of her work wherever they may appear once she has provided a DMCA-compliant notice….[due to 512(m),] the DMCA does not require the policing the Plaintiff suggests.

Secondary Infringement

The court says, without any real discussion, that 512(c) moots Photobucket’s secondary liability. Accord UMG v. Shelter Capital. The court continues with other reasons those claims fail. In particular, Photobucket lacked the requisite scienter about the infringing items it transmitted to Kodak at users’ requests, nor did Photobucket act “in concert” with Kodak. The court rejects the applicability of Grokster, Napster and Aimster because those cases involved peer-to-peer file sharing (more evidence of the exceptionalism towards P2P) and her incomplete takedown notices didn’t confer scienter.


The court granted summary judgment against Wolk, ending her case. The precedential value of this case’s discussion about 512(c) probably will be overwritten by the Second Circuit’s Viacom v. YouTube ruling, and a win against a pro se litigant isn’t much in the grand scheme of things. Nevertheless, the ruling reinforces that courts continue to take the 512 safe harbor seriously. In particular, they continue to rebuff copyright owners who don’t send 512(c)(3) takedown notices but still want judicial relief.

Meanwhile, the “volitional conduct” defense appears to be live and well, especially in the Second Circuit, although I’m not sure anyone understands the doctrine’s parameters.