Court Definitively Rejects AFP’s Argument That Posting a Photo to Twitter Grants AFP a License to Freely Use It — AFP v. Morel
[Post by Venkat Balasubramani]
AFP v. Morel, 10 Civ. 02730 (S.D.N.Y. Jan. 14, 2013)
The facts are interesting and highly relevant to the fast-paced world we operate in. I recommend reading them, if for no other reason than as a cautionary tale. In a nutshell, Morel took photos of the Haiti earthquake, uploaded them to Twitpic and posted to Twitter. Someone else took Morel’s photos and uploaded the same photos to their own account. An AFP staffer found the photos on the second individual’s Twitter account, communicated with him, and thought AFP obtained a license. AFP then distributed the photos to Getty and other downstream third parties. Morel was understandably unhappy and started firing off takedown letters and requests. This prompted a preemptive lawsuit by AFP. (Joe Mullin has an excellent recap at Ars and he delves into the messy factual backdrop: “News flash for the media: you can’t sell photos grabbed from Twitter.”)
Highlights from the court’s ruling:
AFP can’t claim the benefit of any license in the Twitter or Twitpic terms: although this may not be the most consequential aspect of this case, this is one that people will focus on, particularly in light of the Instagram TOS controversy. AFP tried to argue that the Twitter terms of service allowed it to freely use content posted to Twitter (& Twitpic). The court previously said nein, and the court again (different judge) definitively closes the door on this argument. There are lots of quotable bits of the discussion, but this one is pretty cut and dry:
These statements [that you “retain your rights to any Content”] would have no meaning if the Twitter TOS allowed third parties to remove the content from Twitter and license it to others without the consent of the copyright holder.
The court also notes that Twitter’s Guidelines:
further underscore that Twitter TOS were not intended to confer a benefit on the world-at-large to remove content from Twitter and commercially distribute it: the Guidelines are replete with suggestions that content should not be disassociated from the Tweets in which they occur.
[emphasis added] AFP raised a related argument that a broad license was necessary; otherwise, Twitter users would be infringing en masse. The court doesn’t buy this argument either, and this was a pretty wacky argument from AFP. I’m surprised they pressed on with this. They didn’t raise any other defenses to liability, so they (along with the Post) are out of luck.
Other defenses to liability: Getty, one of AFP’s co-defendants, raised two other defenses to liability: (1) it was entitled to a DMCA safe harbor, and (2) it did not act volitionally. I have to admit I was a lot more skeptical of these arguments at the outset, but reading the opinion made me change my mind. At minimum, they are “in the ballpark.”
On the DMCA issue, the court asks this strange question of whether the service provider seeking DMCA protection is “doing something useful for a person or company” or providing “a commodity . . . in the form of human effort.” The court raises the related query of whether “an entity that is directly licensing copyrighted material online is . . . a ‘service provider’ [under the DMCA].” The court intimates that an entity claiming DMCA protection must do something other than merely making content available to third parties. I’m not sure where the court distilled all of this from, but ultimately the court says that there’s a factual dispute as to whether Getty was “actively involved” in the licensing (i.e., was the process totally automated) and whether Getty received a financial benefit directly attributable to the allegedly infringing activity.
Getty loses summary judgment on its no-volitional conduct defense for a related reason. The court says that under Second Circuit law, the volitional conduct defense only extends to the right of reproduction, and here Getty allegedly engaged in distribution. The court also says that there’s a factual dispute as to the extent of involvement by Getty employees in the licensing process.
Neither of these defenses are totally foreclosed to Getty, although I’m skeptical that they will carry the day.
Secondary liability: The court says that there’s enough evidence in the record to create a dispute as to whether AFP and Getty can be held liable for contributory infringement. AFP and Getty both tried to say that they were passive participants at best, and merely passed on the photos, but the court says the available evidence indicates a factual dispute about the extent of their involvement. The court also says that they can possibly be held liable merely for making available Morel’s photos to third parties.
Morel can also assert a claim for vicarious liability against Getty (for some reason he didn’t bring this claim until too late against AFP). The court says that there’s a factual dispute as to whether Getty benefits directly from the allegedly infringing materials and also whether it acted quickly enough to remove the photos from its system (its failure to do so may give rise to vicarious liability). Defendants argued that the DMCA codifies vicarious liability standards, and Morel’s path to derivative liability should be through the DMCA, but the court rejects this argument in a footnote (citing the Second Circuit’s ruling in Viacom).
DMCA CMI Claims: Morel made an argument that AFP’s distribution of the photos with incorrect attribution equaled a claim for removal or alteration of copyright management information or the knowing distribution of material with incorrect information. There were predictable factual disputes around how AFP obtained the photos and whether it knew that the photos were Morel’s when it distributed the photos with incorrect attribution. The DMCA claim requires proof of state of mind, and the court says this can’t be resolved at the summary judgment stage. AFP also tried to argue that the credit information was not “copyright management” information because this information is not intended to provide information about ownership, but the court rejects this argument, taking a broad view of what can constitute “copyright management” information (e.g., adding ‘AFP’ or ‘Getty’ in the caption may form the basis of a CMI claim). Ultimately, factual issues preclude judgment in favor of Getty and AFP on Morel’s DMCA CMI claims.
Statutory Damages: Finally, the parties dispute the scope of statutory damages. Morel says he is entitled to a separate award per work/per downstream subscriber. If Getty or AFP are held derivatively liable for the acts of downstream subscribers, Morel says each such subscriber should result in a separate damage award. This would get really ugly for Getty and AFP. In contrast, they argued that Morel is only entitled to one work per defendant (and jointly liable defendants are only hit with a single award). The court sides with Getty and AFP, saying that they’re only liable once per work and the award against them is not multiplied based on their subscribers. The parties have the same dispute about the scope of DMCA CMI damages, and the court again sides with AFP and Getty. Just one award for each instances of removal of CMI.
Yowza. This is an action-packed ruling that covers a ton of different issues. Tough to do justice in a single blog post. Good fun for copyright geeks, people who like issue spotting, and law profs who can repurpose this for a law school exam question!
It’s nice to see confirmation that AFP’s license argument is silly. We’ve always known this, but it’s good to get further judicial affirmation, particularly in light of the Instagram terms of service dust-up. Service providers may include broad license grants in their terms of service, but it takes a lot to say that the service provider can exploit content or IP outside the ecosystem. It takes even more to say that a third party who is a stranger to the relationship can do so. The court’s interpretation of Twitter’s guidelines and terms is sure to give users some comfort that, when a service such as Twitter says “you own your content,” this at least means that third parties cannot come along and exploit it freely.
Getty’s DMCA safe harbor argument is an interesting one. A purely automated system that takes in photos (or content) and licenses it out theoretically should be entitled to DMCA protection (or at least should be considered a service provider). The court cites to other examples of service providers (YouTube; Photobucket) but distinguishes them by saying they are doing something in addition to merely acting as a licensing engine. I wasn’t terribly persuaded–YouTube is making available user generated videos to the public; it just happens to make them available for free. The court seemed uncomfortable with the notion that a turnkey online service could take in content and license it out and yet fall under the DMCA’s definition of service provider. The court could have just focused on the question of whether the material was stored at the direction of a user; if this answer is yes, moved on to other aspects of the DMCA test (that amply allow Morel to argue that Getty is not entitled to DMCA protection). For some reason, the court went off the rails on to a discussion about the precise type of service that must be provided as a threshold matter. I also suspect the facts in this case will make the DMCA safe harbor argument ultimately difficult. The court’s discussion about what is or is not a “service provider” under the DMCA may have been superfluous in light of factual disputes around Getty’s involvement.
[It’s also worth noting that this isn’t a case against the typical online service provider. Much of Getty and AFP’s operations may be conducted online, but there are still humans ultimately pushing the levers. (At least this is my impression from the facts.) This makes me think we shouldn’t read too much into this as a DMCA safe harbor case or a case that deals with derivative liability for purely online service providers (e.g., YouTube).]
The CMI ruling is interesting. This isn’t the first case to adopt a broad definition of “copyright management information,” but courts seem fairly willing to take this approach. (Tip to photographers and content creators: watermarks are your friends.)
The damages ruling is a breath of fresh air for the defendants. I think the law is somewhat mixed and the court could have gone either way (and have treated individual Getty subscribers as supporting additional awards), but it’s the one saving grace to the ruling from Getty’s and AFP’s standpoint. From a doctrinal standpoint, it’s fairly interesting as well. The court was swayed by what it described as the “absurd” outcome under Morel’s suggested damages framework. I don’t know that the court’s ruling leaves Morel totally out in the cold. He still could claim actual damages and ask the jury to sock AFP and the other defendants, pointing to the revenues they generated. AFP and Getty aren’t the most sympathetic defendants in a copyright case, among other things due to their own zealous and sometimes overly aggressive enforcement efforts. Statute of limitations issues notwithstanding, he could also go after other Getty customers. This would get thorny for Getty.
End result: AFP should have looped in the co-defendants and written a check from day one. After furiously litigating the case for almost two years, it’s definitely going to be writing a check to Morel. Maybe the real argument was around the extent of damages available to Morel, and the court’s ruling on this issue will clear up some grey areas around what he can obtain. Either way, there are a ton of interesting issues addressed in the order, but the simple takeaway is something we already knew: stuff that’s posted online is not necessarily fair game for recycling, especially not photos (and particularly if you’re a corporate defendant who argues for aggressive enforcement when your own rights are allegedly infringed). A service provider’s broad terms of service with its users is unlikely to be of much help to any recyclers.