Tom Brady, I Choose You! How the Tweet-Embedding Decision Is Both Disturbing and Intriguing

We Don’t Serve Your Kind Here

Wow, there have been a lot of really interesting copyright cases in just the last two or three weeks! It’s like being a candy store with money, but you’re full. I’ll try to comment on some of them.
Let’s start with the one that really shocked a lot of people. Please put me in the “not shocked” camp. Here’s the opinion: Goldman v. Breibart News. The court (the U.S. District Court for the Southern District of New York) held that you can infringe a photograph that’s contained in a tweet by embedding that tweet in your publicly-available webpage. Maybe you don’t think that’s shocking, but trust me, to copyright lawyers and Internet industry types, it was shocking.
The plaintiff took a photograph of Tom Brady¹For our foreign readers, he’s the quarterback (the most important position) of probably the best-known and most successful American football team, the New England Patriots. He is also from the Bay Area and attended the University of Michigan, so he is not trustworthy. But that’s not relevant here. walking with Danny Ainge²The general manager for the Boston Celtics, a storied American basketball team. then uploaded the photograph to Snapchat. Because there already was a rumor (completely absurd) that Brady was helping the Celtics recruit another basketball player, the photograph went viral and ended up getting tweeted by several different Twitter users.

Another bedrock principle of the Internet. Photo by Scott Ehardt.

Bedrock in the Fred Flintstone Sense

Now, it is a bedrock principle of the Internet that you can’t be liable for copyright infringement involving the World Wide Web unless you control the server that hosted the infringing content. This is known as the “server test.” Like many bedrock principles of Internet law, it’s based on some squirrelly Ninth Circuit⁵The federal appeals court that covers California and many other western states. precedent⁶And like most Ninth Circuit copyright precedent of a certain period, it involved the nude-image site Perfect 10 (“home of the world’s most beautiful natural women”).: Perfect 10 v. Amazon. That case had to do with Google’s image search and how, when you clicked on a thumbnail image, it linked you to the server where image file was actually stored. Google’s servers didn’t store the image files⁷Though it did store the thumbnails, but that was held to be a fair use. Google no longer stores the thumbnails, I don’t think.
I’m going to carefully explain the Ninth Circuit’s reasoning but only because I have to. At issue was the “public display right,” one of the six rights bundled together to form copyright. These are the the things that only a copyright owner can do, and if you want to do them, too, you either need a copyright owner’s permission or a good defense (like fair use). So, if you “display the copyrighted work publicly,” you infringe copyright. The Copyright Act defines “display”: “to show a copy of [the work].” But wait, in the Copyright Act, “copy” is a term of art! And it means a “material object[], in which a work is fixed by any method now known or later developed, and from which the work can be perceived, reproduced or otherwise communicated, either by directly or with the aid of a machine or device.”⁸Not to exhaust you, but “fixed” is also a defined term of art, but it fortunately doesn’t come into play. But sometimes it does.
Let’s put this all together: To infringe the display right, you have to show a “copy,” i.e., the physical item that embodies or contains the “work.” This implies that you actually have the “copy.” If it were a painting, that’d be easy: you couldn’t show the “work” unless you had a “copy.”⁹Note: the “copy” can be the original painting! I know, I know. You get used to it. Think of “copy” in the newspaper sense, not the photocopier sense (the copyright law term for the latter sense is “reproduce”). But a computer file is different: it has to be stored somewhere, and showing the “work” from the “copy” requires an intermediary device (which is allowed under the definition of “copy”). The “work” would actually be storage device—in this case, a hard drive resident on a server somewhere. That means for Google to infringe the display right, it had to have the server’s hard drive in its possession or control. Hence, “server test.”
OK, I know what you’re saying. Congress clearly had things like paintings in mind when it drafted the definition of “display.” Congress didn’t think about a situation where the “work” can be separated from the “copy” (though, it turns out, Congress did think about that possibility when it defined “perform”). To that I say: tough.¹⁰Having said that, courts have ignored the “copy” requirement with the distribution right, finding that downloading a file can be a distribution, even though only a computer file and not the hard drive is distributed. This also drive me nuts. Congress could easily fix this oversight in a way that would offend nobody, but they’re just hoping courts will bail them out. It’s Congress’ job to write statutes so people can understand what’s legal and what’s not. It’s also Congress’ job to update statutes periodically so they stay relevant.

A Very Public Private

Of course, all this meant was that Google wasn’t directly liable for violating the display right. But wouldn’t it be liable for contributory liability (i.e., intentionally assisting someone else’s infringement)? After all, it’s clearly assisting end-users to display the image on their computers, and surely Google intends its end-users to display the images.
Well, this didn’t come up in Perfect 10, but there’s a fatal flaw in basing contributory liability on the end-user’s displaying the image on the end-user’s computer. To show contributory infringement, you have to show that someone else—the end-user, in Perfect 10—was a direct infringer. But the end-user isn’t a direct infringer because, although the end-user is causing the display, the display isn’t public. It takes place on the end-user’s own, private computer.¹¹In Perfect 10, Perfect 10 didn’t base its claim of contributory infringement on the end-user’s conduct, but on the websites to which Google was linking. They had illegal copies of Perfect 10’s images. Unfortunately, the Ninth Circuit didn’t have enough information to determine whether Google materially contributed to those sites’ infringement and remanded that question for the trial court. I cannot tell what happened to that claim after remand.
For a lot of people, Perfect 10 looked like a hyper-technical reading of the Copyright Act that allowed Google to get away with massive infringement of the display right. Others acknowledged that it was technical, but then the Copyright Act was a pretty technical statute. It would hardly be the first counter-intuitive result in copyright law, and at least we knew were we stood. It would be up to Congress to work this out. Indeed, sometimes the technical nature of copyright law leads to the finding of infringement in unexpected ways, like when you take your Kindle across national borders.
Edited to add: Mark Whipple believes the “Server Test” makes no sense, and his reasoning is better than the Court’s in Goldman. His point is that you don’t need to possess the copy, just have visual access to it. Video transmission to the public of a painting (private or public) would still implicate the display right. I don’t “have” the painting, and I’m not even in the audience viewing my own transmission. I’m just doing the transmitting. To him, there’s no great difference between this kind of physical access to a copy, and computer access to a remote image file. I think under this view, we don’t need to worry about “volitional conduct” (about which more later).

Poison Aereo

Then along came the Supreme Court’s deeply odd decision in American Broadcasting v. Aereo. Remember that case? It involved a way of sharing broadcast (over-the-air) TV signals over the internet. The defendant, Aereo, used hundreds of separate TV antennas, each of which was being (essentially) rented by a particular subscriber. I blogged extensively about this, but the key blog post is this one. My point was that, if we treated copyright law as a purely technical law, Aereo should have gotten away with its crazy scheme because each antenna “belonged” to an individual subscriber, who caused the TV program to “perform” privately. No direct infringer, no contributory infringement. Easy-peasy.
Only this was a bridge too far for the Supreme Court, and it ruled against Aereo. The Supreme Court’s reasoning is… fuzzy. Instead of fitting its decision into the body of copyright law, it basically just said that Aereo resembled a cable company, and cable companies publicly performed the works they re-transmitted. The dissent, led by those who prefer statutes to say what they mean¹²I.e., Scalia, Thomas and Alito. On the plus side, the case got Ginsburg and Breyer on the same side of a copyright dispute, which is something. would have let Aereo off the hook because, in their view, copyright law was technical law, and the majority wasn’t respecting that. It was left to copyright lawyers and scholars to figure out what, if anything, we could take from Aereo.
It was hard to know what to do with Aereo. Was it an anomalous case, destined to be “limited to the facts”? Was there some broader lesson to learn about what it means to be a “direct infringer,” and especially the concept of “volitional conduct”? “Volitional conduct” was supposed to help us identify the direct infringer, especially in cases involving the public display and public performance rights, where it matters whether the direct “infringer” is a public or private person (because the private person isn’t an “infringer” at all). But all it really means is: the party responsible for causing the performance or display.
Until Aereo, the leading answer to that question was: whoever pushed the button. Thus, the dissent in Aereo argued vehemently that the subscriber of the Aereo service was the responsible party because he or she is the one who pushed the button on the remote that caused the work to be performed on the TV or computer.¹³The majority in Aereo doesn’t mention “volitional conduct” at all, whereas the dissent doesn’t shut up about it. But that wasn’t the only choice. Lots of people preferred the idea of “proximate cause.” This means not only did you somehow cause the bad thing that happened (in some chain of events), but you can kind of see the bad thing from the thing you did (a concept known as foreseeability). Yes, this is a very squishy concept, especially when compared with “who pushed the button,” and you need a jury to make the call. But it is also a concept very familiar to lawyers and judges. We can’t describe it, but we kind of know it when we see it.

I Looked Under the Hood but I Couldn’t Even Find the Carburetor

But the judge in Goldman drew a different lesson, one Mike Masnick worried would come to pass: copyright “liability should not hinge on invisible, technical processes imperceptible to the viewer.” Under this view, we don’t care about what’s happening “under the hood,” just how things seem to the ordinary observer.
If that sounds like some big win for copyright holders, think again. Copyright protection of software and digital content relies heavily on what goes on “under the hood.” You need a license to use software and digital content that you purchased because, to use them, your computer/Kindle/iPad/whatever has to make a temporary copy (er, reproduction) into your device’s memory (from storage), and that implicates the reproduction right. But if someone didn’t tell you that, you couldn’t have been able to tell just by looking at it. The ReDigi case, which held you can’t re-sell digital content the way you can, say, a music CD, would have to be reversed under this view. You can’t see the copying (er, reproducing¹⁴Wow, that sounds weird.). As far as you’re concerned, your transaction looked like re-selling a CD. Copyright law is often surprising. And copyright holders might wish to remember that the first to get unpleasant surprises were consumers.
The court went on to trash the “server test” as somehow not grounded in the text of the Copyright Act. Since the court is in New York, it doesn’t need to accept the Ninth Circuit’s holding in Perfect 10. It can assess the Ninth Circuit’s reasoning on its own merits.¹⁵That’s why where this case was filed is sort of important. The court doesn’t actually get into the weeds of how the Ninth Circuit actually arrived at the “server test”—you know, looking at the relevant statutory language, because Aereo taught us that looking at the relevant statutory language is for chumps. Instead, it looked at a statutory defense to infringement of the public performance and public display rights: the right to retransmit a TV signal (i.e., receive a TV signal then send it along via an antenna, cable, etc.). The court noted that the defense doesn’t contemplate possession or control of the “copy” of the work being transmitted. In an analogue world, you don’t make a copy of the work in order to retransmit; you just convert the signal and re-send it. Do you know why this defense doesn’t contemplate having a copy of the work? Because the definition of “performance” was carefully crafted not to require one. By contrast, as we’ve seen, the definition of “display” does require having a copy of the work.¹⁶Why is the display right even included then? First of all, it’s not unusual for affirmative defenses to be a bit over-inclusive because they’re often trying to account for possibilities rather than certainties, and there’s no harm in it. I.e., if courts found X to be infringing, then at least we’ll know that X has an affirmative defense. It’s easier than explicitly carving out X. They also guard against future changes to the statute itself. If the display right were to be re-defined later, we won’t have to update this defense to include it.

Choosey, Choosey

The court then distinguished Perfect 10 on grounds that, in Google’s case, the end-users were the ones choosing which images to display, whereas in this case, the website operators did the choosing (and it found some useful language in the dissent to Aereo for this proposition). And that, my friends is a very, very…
…intriguing point. Look what it does to our concept of “volitional conduct.” It’s a compromise between the hyper-technicality of “who pushed the button” and the squishiness of proximate causation. I think it would be fairly easy to apply. I suppose there’s some room between no choice at all (the case here) and lots and lots of choice (the case in Aereo). What if the reader/viewer had only two choices? But I don’t immediately see some clever way around it in the context of tweet-embedding (e.g., by offering the reader a choice of tweets to view: the tweet behind door no. 1 or the tweet behind door no. 2).
Perhaps the real difference is not so much choice between programs, as a choice to view the copyrighted work at all. When the reader of the online journals in Goldberg clicks on a link to the webpage, the reader only knows she’ll be reading about the absurd Tom Brady-Celtics rumor, not necessarily that she’ll be viewing the plaintiff’s photograph.¹⁷Though she probably expected to see SOME photograph, she didn’t choose to see that or any photograph. But when one subscribes to Aereo (or uses Google’s image search) the subscriber is making a deliberate choice to watch the copyrighted work at issue.
Anyway, this is worth exploring. What isn’t worth exploring is the idea that we can’t look under the hood at what the wires, memory chips, microprocessors, bits and radio signals are doing when we analyze copyright infringement. There are many ways to look at copyright law, but “heads I win, tails you lose” isn’t one of them. If we look under the hood when it benefits copyright owners or consumers (depending on which “side” you take), then you have to look under the hood when it hurts copyright owners or consumers.
And while we’re at it, yes, it would be nice if copyright law were treated as an internally consistent body of law by the Supreme Court, but we’ll endure.
Thanks for reading!

Rick Sanders

Rick is an intellectual-property litigator. He handles lawsuits, arbitrations, emergency injunctions and temporary restraining orders, opposition and cancellation proceedings, uniform dispute resolution proceedings (UDRPs), pre-litigation counseling, litigation avoidance, and other disputes, relating to copyrights, trademarks, trade secrets, domain names, technology and intellectual-property licenses, and various privacy rights. He has taught Copyright Law at Vanderbilt University Law School. He co-founded Aaron | Sanders with Tara Aaron-Stelluto in 2011.