Richard Posner LOLs a Bad Preliminary Injunction. But Does Anyone Get the Joke?
Is Richard Posner’s opinion in Flava Works v. Gunter (better known as the “myVidster” case) already a month old? For such an anticipated decision, it sure flew under the radar after a little bit of buzz. It’s already been pronounced something of a dud*, which is perceptive, if not entirely fair.
The decision was so anticipated because it was appealed from a lower-court opinion that, according to popular (but slightly erroneous) report, implied that merely linking to unauthorized material was an act of infringement (though there might be a DMCA safe harbor for that). The truth was only slightly less alarming: the lower-court decision held that merely framing content that is hosted elsewhere is an act of infringement.
To add to the drama, the appeal was heard by a panel that included one of America’s best known and highly regarded jurists, Richard Posner, whose copyright opinions were known to be somewhat erratic. Oh, and the copyrighted material in question? Hardcore porn*. That gets everyone’s attention.
* If it weren’t for porn, how much less would copyright and internet law be developed?
The defendant, myVidster, ran a kind of social bookmarking site. If you find video content on the web, you bookmark it at myVidster’s site, which then obtains the video file’s “embed code” and incorporates that code into its own html code. The result is familiar to anyone who uses the web: a little mini-player that appears to be on myVister’s site—this is known as “framing”—plays the video. However, the video is not hosted by myVidster. The rest of the necessary “equipment”—the browser, the player, and any necessary plug-ins—is actually supplied by the user (whether the user is all that conscious of it or not). All myVidster does is supply the site and the html code. But it looks like myVidster is doing everything.
The plaintiff, Flava Works, sued and sought a preliminary injunction. I describe the general contours here, but the salient point is that hearings on motions for preliminary injunction can be rushed, messy affairs under the best of circumstances. It’s a bit like having the trial at the beginning of the case, before all the facts have been discovered, before the legal theories have been developed, and before the lawyers have their heads wrapped around the issues. When the preliminary injunction involves complex copyright law and poorly understood technology (and, wow, was it poorly understood*), the result might be a train wreck.
* myVidster’s proprietor didn’t actually know how his site worked. This caused Flava to focus on the wrong aspects of myVidster’s technology, and to make myVidster come off as evasive at the hearing.
The other salient point is that, precisely because it’s so messy, orders granting or denying preliminary injunctions are given a great deal of deference. The trial court is basically asked to make a gut-level decision, one that can be devastating to the loser if it’s wrong, and the appellate court usually respects that. In short, preliminary injunctions are rarely overturned.
So it should come as no surprise that the trial court’s opinion is pretty much completely bollixed up. To make a claim for contributory infringement, Flava had to prove (1) a third party infringed Flava’s copyrights, (2) myVidster knew about the infringement, and (3) myVidster “materially contributed” to the infringement. Now, third parties had somehow acquired Flava’s copyrighted A/V files and uploaded them to third-party sites. There was no evidence that myVidster had anything to do with those uploads, but still, that’s the act of infringement the court latched onto.*
* What about the act of playing the videos? Wouldn’t that make for a better case of contributory infringement? Um, yeah, that’s a really, really good question….
How did myVidster know about the infringement? Why, Flava told it so, through DMCA take-down notices. Now everyone gets really confused because the focus shifts to the DMCA safe-harbor defense, before infringement has been established. But let’s step back: are DMCA take-down notices evidence of knowledge, when they come well after the fact? Sure, myVidster can remove the embed code from its html code, but the embed code isn’t the act of infringement at issue—it’s the uploading of the A/V computer file to an unrelated site, which could have happened years ago for all anyone knows.
So, how did myVidster materially contribute to those infringing uploads? The judge had five reasons:
- The judge blamed myVidster’s existence for those uploads. He assumed that the uploads were done by myVidster’s own users, though that’s not necessarily the case. It’s probably a safe assumption, but there was no evidence of that. Anyway, the uploads happen before they are “bookmarked” on myVidster’s site. It’s possible some myVidster user or users illegally uploaded Flava’s A/V files simply so they could bookmark them on myVidster, but (a) that doesn’t make a tremendous amount of intuitive sense, and (b) there was no evidence that ever happened.
- The judge blamed myVidster’s “video backup” service. You see, the problem with letting third parties host the content is you can’t always be sure the content will be there when you want to view it. So myVidster had a service that let you place a copy of the A/V file on myVidster’s site. That’s clearly copyright infringement—whether direct or contributory on myVidster’s part is academic. But, um, that’s a different act of infringement than the illegal uploads that we’re actually talking about, isn’t it?* Gotta keep you eye on the ball.
- The judge blamed myVidster’s social-networking “sharing” feature. But remember: all that is “shared” is a link to the content. That happens well after the fact of the illegal upload. The judge seemed to think that “sharing” meant copying or something.
- The judge blamed myVidster’s failure to warn its users not to link to infringing content. But that begs the question of whether linking or framing is itself illegal.
- Finally, the judge blamed myVidster’s failure to have and implement a repeat-infringer policy. But we’re getting ahead of ourselves. That’s an element for the DMCA safe-harbor defense. It’s not an element for infringement. Just because you don’t fit within one of the DMCA safe harbors doesn’t mean that you’re an infringer. It just means that, if you are an infringer, you’re screwed.
* In fact, the injunction against this practice (which myVidster had stopped, apparently, because of: illegal) remained in effect.
So, yeah, the lower court really blew it. One really gets the sense that it had lost all patience with myVidster and wanted to find some way to hammer it, and Flava took advantage of that and provided the judge with the easiest route. But the easiest route isn’t always the correct one.
The pity of it is that a much stronger case could have been made if Flava had focused on the act of playing the A/V files. Why, exactly, was everyone focused on the illegal uploads? Playing the files was just as illegal*! It’s not a slam dunk by any means, but a much stronger and more intuitive case. Oh, and what about vicarious liability? Well, that wasn’t argued, either.
* The performance right is the main exclusive right implicated here, but non-live streaming, more technically known as “progressive downloading,” may also violate the reproduction right, depending on the size of the buffer and the fate of the already-viewed material. You may read my agonizing about this issue here.
Play the Hand You Were Dealt. Or Not.
So, this is what Judge Posner, writing for the Seventh Circuit, had to work with. Yes, there are much more intuitive and sensible ways to approach the problem, but those ways weren’t available to Judge Posner—not that it stopped him from trying (as we’ll see)!
For all the deference appellate courts are supposed to give trial judges in this situation, Judge Posner’s decision isn’t a very hard one—the lower court decision was practically begging for reversal. For one thing, the trial court applied the incorrect standard for preliminary injunctions. In ages past, courts could presume irreparable harm, but eBay changed all that. This is hardly a secret.* Judge Posner could have vacated the preliminary injunction** right there.
* eBay was a patent decision, but it’s pretty well known that it applies to copyright injunctions, too.
** So we’re clear: that’s the injunction as it applies to myVidster’s “framing” the A/V content, not to its “video backup” service. Everyone agrees that should be enjoined.
Still, you can’t blame Judge Posner for not stopping there. If you exclude all of his dicta (stuff that doesn’t count as precedent because it isn’t necessary to decide the issues actually before the court), his next holding is that myVidster didn’t contribute to the illegal uploads. He based this decision on the remoteness of the uploads to anything myVidster did, applying, as he said, “general tort principles.”*
* He mocks Google’s and Facebook’s amici brief on this point. (You can imagine why they’d be siding with myVidster!) They try to describe myVidster as a “tertiary infringer,” with the bookmarking users as “secondary infringers.” Posner describes this legal theory as “a species of layer cake.”
Then Judge Posner goes on a long, fairly thoughtful, but strange digression about the performance rights at issue. The parties wanted to debate whether the “server test” from the Ninth Circuit’s Perfect 10 v. Amazon decision, which (despite its name) involved Google’s display of Perfect 10’s pictures*. Perfect 10 is pretty much “on point” here, if you swap “display” for “performance” (which makes no difference to the analysis). The Ninth Circuit held that Google wasn’t liable because it didn’t host the content, so it couldn’t “display” it.*
* Of nude models. Seriously, where would copyright law be without porn?
** Shockingly, the trial court ignored the whole issue. So why does myVidster raise it on appeal? I assume it was raised by Flava at the preliminary injunction hearing. Right?
But Posner was interested in more fundamental question, one that the parties didn’t brief (and couldn’t have known to brief). When does a public performance occur? To his way of thinking, the Perfect 10 decision assumed that a public performance occurred when the content was poised and ready to be played. And Posner appears to accept Perfect 10’s conclusion, if the assumption were true.
But would the analysis be different if it occurred when the user pressed “play”? You have to admit that, in that case, myVidster is making the infringing public performance awfully easy for the user (which would be the direct infringer). The analogy there would be a swap meet that the organizer knew was being used to sell pirated music, which has been held (again by the Ninth Circuit) to constitute contributory infringement on a “support services” theory. Posner decides the analogy doesn’t quite hold because, unlike the swap meet organizer, myVidster doesn’t take a cut of the proceeds.*
* Doesn’t that sound a lot more like vicarious liability? And is vicarious liability specifically NOT before the court? Argh!
Here’s the thing: all of this analysis is basically Posner arguing with himself because the parties didn’t brief these issues at all, and Posner is coming at the issue from an unexpected (but interesting) direction.* Maybe Posner should have asked the parties to re-argue this part of the case, but why bother? The preliminary injunction was so bad, it was going to be reversed no matter what.
* Doesn’t that make it all dicta? Well, not really, since the analysis was necessary to the holding. It’s just that the judge’s reasoning wasn’t tested by the adversarial process.
By the way, Posner essentially followed Cablevision here in assuming that the user was the direct infringer. (IMPORTANT EDIT: No, this isn’t quite right. Cablevision held the customers were the direct infringers for the copying, but avoided deciding whether they were also the direct infringers on the public performance. In theory, the analysis should be the same, but you never know. Hat-tip to David Hartline.) His layer-cake crack notwithstanding, even Judge Posner recognizes that you have to distinguish direct infringers from indirect infringers (what the rest of us would call “secondary infringers”) because of the knowledge requirement that’s missing from indirect infringement.
I’ll have some further thoughts about this strange little opinion next time .
Thanks for reading!