Nieman Foundation at Harvard
The News Lens in Taiwan is doing what media startups in the region hesitate to do — acquiring other sites
ABOUT                    SUBSCRIBE
Feb. 16, 2018, 2:21 p.m.
Reporting & Production

Are news publishers directly liable for embedding tweets that contain images not created by that tweeter?

A New York federal judge ruled that when publishers from The Boston Globe to Vox Media to Breitbart “caused the embedded tweets to appear on their websites, their actions violated plaintiff’s exclusive display right.”

News publishers will now be thinking twice before embedding in articles a certain viral tweet that contains a “full color image” of Tom Brady and Danny Ainge, taken in 2016. In fact, any publisher embedding any tweets containing content that might not be original to the tweeter now has at least some reason to be concerned.

On Thursday night, a New York federal judge ruled that simply embedding certain tweets can constitute a copyright violation by news publishers. The judge wrote (via The Hollywood Reporter):

Having carefully considered the embedding issue, this Court concludes, for the reasons discussed below, that when defendants caused the embedded Tweets to appear on their websites, their actions violated plaintiff’s exclusive display right; the fact that the image was hosted on a server owned and operated by an unrelated third party (Twitter) does not shield them from this result.

The photo in question, taken by a man named Justin Goldman, was newsworthy at the time. It showed that the New England Patriots quarterback was helping the Boston Celtics in its attempt to lure free agent Kevin Durant. If you want to see it, it’s still up at NBC Sports, at least at this writing. A Google image search also works. Here’s one of the tweets referenced in the lawsuit, from Oklahoma City Thunder reporter Travis Singleton. Note that the image “has been removed in response to a report from the copyright holder.”

The defendants in the case: Breitbart, Heavy Inc., Time Inc., Yahoo, Vox Media, Gannett, Herald Media, Boston Globe Media Partners, and New England Sports Network. You can read the entire decision here.

The collective ears of publishers across the country perked up at the unexpected decision.

I spoke with Andy Sellars, the director of the Technology and Cyberlaw Clinic at Boston University School of Law, who offered some clarifying responses to my extremely basic and frantic questions. We talked about how the “jargon of the internet and internet culture is very different from the jargon of the law,” the confusion that results from the meshing of the two languages, and how long it may be before we get an official resolution on the case.

Shan Wang: I’m freaking out because I’ve definitely embedded tweets that contain content that may not have been original to the tweeter. Also, I tried to read the full opinion and admit I have no idea what’s actually going on.

Andy Sellars: There’s direct copyright infringement and contributory copyright infringement. Direct is the actual reproduction of an image; copyright is the right to make copies, right? Then there’s contributory infringement, which is when you’re not the person who does the infringement, but you’re aware of the infringement and then you somehow contribute to it. This is something web companies worry about because they provide platforms, and there’s always the possibility that a copyright owner can say, you knew this platform was being used to infringe on my work, and you contributed to it by providing a platform.

We have in the area of contributory liability two important limitations. It requires you to be aware of it, whereas direct copyright infringement — I don’t even have to be aware that I’m infringing copyright, to be liable.

The other thing in the context of contributory copyright infringement, we have what’s known as the DMCA “safe harbors,” which protect platforms like Google and Twitter when they link to other people’s content or host user-generated content.

Distinguishing between direct and contributory copyright infringement is something that used to play in the early days of internet courts, but was kind of settled more or less in the Kelly v. Arriba Soft Corp. case. Pretty much every court since then has adopted what’s known as the server test. Which basically says I’m not going to look at how something is presented on the web, I’m going to look at where the particular image or text or whatever the infringing thing is hosted. The owner of that server is the one directly infringing, everyone else is at best contributory infringing.

What this case does is that it goes back to the old way of looking at it: We don’t care as much about the server test, we just care about the basic effect of what the website looks like.

You can host content on the web either by uploading to your own server, or by linking someone or framing someone’s content that’s hosted on another server. Now you’re not just looking at where the object comes from and whether that person has handled all the copyright questions — anyone who then displays the object could be responsible.

In this particular case, it’s still a little early. It’s not set in stone and the case will almost certainly be appealed. But the suggestion the court makes here, is that The Boston Globe and others, by embedding a tweet, can be held directly liable for infringement of content in the tweet. That makes it a lot scarier to embed tweets going forward.

Wang: Here’s what I imagine publishers really want to know. It doesn’t mean that at the moment, if I embed a tweet that isn’t mine, I will be in huge legal trouble?

Sellars: This is a pretty remarkable finding already. I doubt the plaintiff’s lawyers are super eager to push it.

So for example: When I take a photo of my dog and upload that to Twitter, I’ve given Twitter a license to use that photo in certain ways. That license extends to using the embed feature on other platforms. If I’m confident that the person who posted it to Twitter is the person who took the photograph, we’re okay, because we have this license argument we can use.

Where these publishers got into trouble in this case was in the fact that the person who posted the tweet that had the image was not the photographer.

Wang: That’s…everything on Twitter?

Sellars: There’s some things that are obviously not; memes, reactions, and so forth. In this particular case it was a little bit hard to tell.

It used to be you would embed a tweet and not worry about clearing your rights to the underlying image. If there was an issue, they’d probably go to Twitter to work it out. If it got taken down under a DMCA notice, then the only thing that would happen for you is that the tweet wouldn’t load.

Whereas now, this suggests you could still be liable even though you didn’t know necessarily that the image wasn’t authorized to be posted to Twitter, because the court says you are the one displaying it, and not Twitter.

That’s why, at least, all the copyright lawyers in the world are standing up and paying attention this morning, because that’s a departure from the way we thought about copyright on the web for over 15 years now.

Wang: The other thing that is confusing and scary is that there are terms used in the ruling that are so…strange that I’m not even sure what they refer to. Weird ways to describe to what a Twitter embed actually is, technically speaking, for instance.

Sellars: That’s the hard thing I deal with all the time. The jargon of the Internet and Internet culture is very different from the jargon of the law. And they both have a lot of jargon.

When we’ve been too metaphorical in the past about how the Internet works and rely on analogies that a court might be more familiar with, it means courts can use more dangerous inferences to say, well, your website displays the image, therefore you’re liable under the display right copyright law. Rather than saying, you’re not displaying the image, what you’re actually doing is calling the third-party server that has the image, and it’s the third-party server that needs to assess the copyright risk, and you would only be liable after you’ve been notified that there’s been an infringement issue.

There’s been some thought about this. Orrin Kerr, a computer scholar, had this article called the Perspective of Internet Law. He addresses this: Depending on which analogies you use to describe the web, you’re also going to lead courts to very different results. If you’re looking at the impression of a website, it’ll lead you down one path. If you’re looking at how the website is coded and loaded, it leads you down a very different path.

These are the sorts of problems Internet lawyers have to deal with. Which one is the more appropriate way to analyze things?

Wang: What can we impatient journalists expect for a timeline? Since the case may be appealed and it’d be great to know what exactly we’re allowed to publish online.

Sellars: I haven’t dug into this too deeply yet, but I think this was a denial of a partial summary judgment. So that means there either will be further discovery or there will be a trial. I wouldn’t be surprised in this case if the media companies ask the court to make an interlocutory appeal, which is to say, normally, you wait to appeal until you’ve reached the end of the case. In this case they may say, your honor, this is important, let’s pause this case, appeal this to the federal appeals court, and have them weigh in on this direct liability versus contributory liability question, before we continue with the case.

This strikes me as one of those cases where this may very well happen. And the court will either accept or reject it, and if they reject it, you wait until the case is totally done, which either means there’s going to be a trial, or some sort of stipulated judgment. So it could be a while.

When it goes to the appeals court, it’ll still be a while. It could be several months up to a year or perhaps even longer. There’s another case that’s pending before the court right now involving these programs that systematically monitor television networks and create a searchable index of what’s said on TV. That’s been before the court almost two years now. So it could be a while.

Wang: Why did the plaintiff file this in New York? I’m wondering if there are more favorable precedents for this kind of thing in New York versus Massachusetts where a lot of the offending news organizations like the Globe and Herald are based.

Sellars: Usually a plaintiff decides where to file it, and the only restriction is that it has to be in an area where it’s considered fair as a matter of due process to bring the defendant in. Most of the defendants in this case are big media companies, so it’s fair to bring them to pretty much any court in the U.S. It probably was convenient for the plaintiff to bring them to New York, versus Massachusetts, where a lot of the defendants like the Globe and the Herald are based.

Often there’s a strategic calculation about where to first bring a case, but I’m not aware…I guess I could make an argument that questions are a little more restrictively read in New York than in Massachusetts. We have some good fair use cases in Massachusetts that help us out here. I could see a potential argument here. But it probably has more to do with where this would need to be for the person to actually show up at court and be able to do stuff.

POSTED     Feb. 16, 2018, 2:21 p.m.
SEE MORE ON Reporting & Production
Join the 45,000 who get the freshest future-of-journalism news in our daily email.
The News Lens in Taiwan is doing what media startups in the region hesitate to do — acquiring other sites
Within the first three months of this year, the four-year-old news company brought a tech site and a top site for sports fans under its umbrella in order to grow its overall reach: “By the end of this year, I want to have 5 or 6 different sub-brands,” News Lens co-founder Joey Chung said.
Could students’ media literacy be compared across countries, like math scores?
Plus: YouTube’s sad reliance on Wikipedia, a problem with Pinterest, and how countries around the world are fighting misinformation.
The Join the Beat project wants to tease out better ways of working with an audience directly and regularly on stories
“We want to find people who recognize the potential of the audience, and are looking to capitalize on this potential and pull together a learning group to try some things and see what emerges.”