Copyright Trolls Argue Embedding Videos Kills Innovation and Profits.
MPAA: you can infringe copyright just by embedding a video The Motion Picture Association of America is squaring off against a coalition of Internet giants and public interest groups over the key question of whether it's possible to directly infringe copyright by embedding an image or video hosted by a third party.
A federal judge
took that position last July, prompting a chorus of criticism. Two briefs-one by Google and Facebook, the other by the Electronic Frontier Foundation and Public Knowledge-attacked the decision as contrary to past precedents and potentially disruptive to the Internet economy. They asked the Seventh Circuit Court of Appeals to overturn it.
Last week, the MPAA joined the fray with a brief in support of Illinois federal judge John F. Grady's ruling. It urged the Seventh Circuit not to draw a legal distinction between hosting content and embedding it. In the MPAA's view, both actions should carry the risk of liability for direct copyright infringement.
The case arose from a dispute over Internet pornography. MyVidster is a video bookmarking site that allows users to save links to their favorite videos and share them with others. The site supports embedding, so bookmarked videos can be viewed on a myVidster page surrounded by myVidster ads.
A porn company called Flava Works discovered numerous myVidster pages with embedded Flava videos. MyVidster did not host the videos-they were streamed by third-party hosting sites such as RedTube-but Flava sent myVidster DMCA takedown notices anyway. MyVidster claims it complied with these takedown notices, but evidently its response didn't satisfy Flava, which eventually sued for copyright infringement.
Hosting vs. embedding
The case has been plagued by confusion over the difference between hosting a video and embedding a video hosted by someone else. MyVidster's owner, Marques Gunter, failed to clearly explain the distinction in his testimony, and Flava's lawyers exploited Gunter's poor communications skills to make him look evasive and indifferent to following the law.
For example, a key issue in the case is whether myVidster qualifies for a safe harbor under the Digital Millennium Copyright Act (DMCA). To qualify for the safe harbor, a service provider is required to have a policy of terminating the accounts of repeat copyright infringers. Gunter apparently interpreted the DMCA as only requiring him to terminate users who directly infringe copyright, and he believes that bookmarking (and, consequently, embedding) a video does not qualify as direct infringement. Hence, he didn't terminate users who bookmarked publicly available videos, even if those videos infringed copyright.
Reasonable people can dispute Gunter's interpretations of copyright law (indeed, the MPAA does just that) but it's not clear that Judge Grady even understood Gunter's position. Gunter testified that he would warn, and eventually terminate, the accounts of users who "use myVidster to publish links/embeds of videos that would otherwise not be accessible by the public. For example, if a user is uploading videos to a file server and using myVidster as a way and means to distribute the content."
However, Gunter said that "most of the content are embeds which are hosted on external websites, [and] I would suggest contacting the websites that are hosting your content to help stop the future bookmarking of it on myVidster."
Judge Grady characterizes this as "the epitome of willful blindness," writing that Gunter "pointed a finger at other websites while failing to acknowledge that his own website is perpetuating copyright infringement." But Gunter's point wasn't that infringement on other sites excused the infringement of his own users. It was that embedding an infringing video wasn't copyright infringement for purposes of the DMCA's "repeat infringer" rule.
Grady wasn't impressed with this distinction-indeed, he seemed to regard it as so obviously wrong that he doesn't even bother to analyze it in any depth. Instead, he ruled that because the infringing videos appeared "on myVidster," myVidster was responsible for them regardless of the technical details of how they were delivered to the user's computer.
But as Google and Facebook pointed out in a November brief, previous precedents have found the distinction between hosting and embedding legally relevant. The
2007 decision of Perfect 10 v. Amazon is a key precedent. The Ninth Circuit Court of Appeals held in that case that only the server that hosts infringing content, not other servers that may provide links to infringing content, are guilty of direct copyright infringement. The Perfect 10 decision is a strong precedent because it dealt with the closely analogous case of Google's image search engine embedding copies of infringing images hosted on third-party servers.
Grady also ruled that Gunter's failure to take proactive steps to discourage infringement by myVidster's users undermined his claim to the safe harbor. Grady complained about Gunter's "mechanical response" to takedown notices, faulting him for failing to "warn his users about copyright infringement," to prevent previously blocked videos from being uploaded again, or to terminate users who have posted infringing videos "on two or more occasions."
The MPAA weighs in
Judge Grady's decision was music to the ears of the MPAA, which has long downplayed the distinction between websites that host content and websites that merely link or embed content hosted elsewhere. Last Wednesday, the MPAA filed an amicus brief urging the Seventh Circuit to uphold key parts of Judge Grady's decision.
"Although there is nothing inherently insidious about embedded links, this technique is very commonly used to operate infringing internet video sites," the organization writes. "Pirate sites can offer extensive libraries of popular copyrighted content without any hosting costs to store content, bandwidth costs to deliver the content, and of course licensing costs to legitimately acquire the content." The MPAA also notes that embedding can enable sites to monetize infringing content by surrounding it with ads.
The MPAA argues that "myVidster users who posted links to infringing videos and images participated in the process by which those videos were streamed and shown to the public," making them direct infringers. It calls the "server test" articulated by the Ninth Circuit in the Perfect 10 case "novel," arguing that it "conflicts with the language and legislative history of the Copyright Act, which affords copyright owners a broad and technologically neutral display right."
In the MPAA's view, then, anyone who serves a website with an embedded link to infringing content "displays" that content to users, directly infringing copyright in the process. The MPAA also backs Grady's narrow conception of the DMCA's safe harbor. It argues that "the safe harbors only apply to 'innocent' service providers." And myVidster, in the MPAA's view, doesn't make the cut.
Far-reaching implications
If the Seventh Circuit adopts Judge Grady's-and the MPAA's-expansive interpretation of copyright liability, implications for the Internet economy could be far-reaching.
Numerous websites embed content from third parties they have not personally inspected. Under the theory articulated by Grady, and supported by the MPAA, these websites would be responsible for this content, exactly as if they had stored it on their own servers. This could create a serious disincentive for sites to allow users to post embedded content, hampering the convenience and user-friendliness of the Web.
Of course, as Google and Facebook stress in their brief, sites like myVidster could still be liable under the secondary infringement rules that felled Napster and Grokster. So the argument that myVidster's users are not a direct infringer isn't an argument that "link sites" that profit from infringement should get a free pass.
Secondary infringement is more difficult to prove than direct infringement, however, and the penalties are lower. That's as it should be. Sites that host content should bear greater responsibility for that content than sites that merely link to content hosted by third parties.
The Lawsuit That Could Change Video Embedding As We Know It Over at Ars Technica,
Tim Lee brings news of a disturbing lawsuit, now supported by the Motion Picture Association of America, that could set a legal precedent that embedding copyright video, rather than hosting it, counts as copyright infringement:
“Although there is nothing inherently insidious about embedded links, this technique is very commonly used to operate infringing internet video sites,” the organization writes. “Pirate sites can offer extensive libraries of popular copyrighted content without any hosting costs to store content, bandwidth costs to deliver the content, and of course licensing costs to legitimately acquire the content.” The MPAA also notes that embedding can enable sites to monetize infringing content by surrounding it with ads…
Numerous websites embed content from third parties they have not personally inspected. Under the theory articulated by Grady, and supported by the MPAA, these websites would be responsible for this content, exactly as if they had stored it on their own servers. This could create a serious disincentive for sites to allow users to post embedded content, hampering the convenience and user-friendliness of the Web
I, and the rest of my colleagues at ThinkProgress (not to mention our peers elsewhere on the internet), would have to dramatically reassess the way we do business, were this precedent to become law. Embedding is an elegant tool for journalists, and a great convenience for readers. It lets us write posts and stories that have a neat flow to them, framing a piece of content, letting the reader consume that content, and then move on to our analysis all without forcing them to click away, perhaps never to return. Sure, it keeps people on our site and lets us make money, but it’s also a convenience for the reader that provides a coherent consumption experience.
If this legal precedent is established, it would create a hugely complex situation. There’s a lot of content that the copyright holders would like to see widely embedded and distributed, whether it’s move trailers, music videos, campaign ads that no one actually intends to spend money to air but they would like to be seen, speeches, etc. That desire isn’t going to disappear if a new legal regime governing embeds comes in place. And that creates a terrific problem for both people who want their content embedded and those of us who need to embed a wide variety of content to do our jobs. Given the huge amount of content out there, and the large number of vectors through which it’s made available, it’ll be extremely difficult to comply with a new regime if there’s no clear way to tell if the content’s licit or not. And without that clarity, media outlets might be less willing to distribute even licit content if they can’t clearly document its provenance. That skittishness could prevent transmedia campaigns like the Peter Weyland TED Talk that’s being used to promote Ridley Scott’s Prometheus, where video is meant to appear sui generis rather than clearly coming from a studio, from taking off, which would be a loss both for the content producers themselves, and for the people who would enjoy that content if it was distributed to them.
Given the fuzzy, burdensome precedent this lawsuit could set, I’d like to see the clear numbers that explain why the potential use of embedding for intentional copyright violation is so harmful that it justifies upending the legitimate use of embedding for the rest of us.