Goldman, Breitbart and embedded tweets: another link in the copyright chain

Adam BushnellPosted by

This article looks at the continuing evolution of the law of copyright in the US, and considers how US case law might influence lawyers in the EU.

The story begins in July 2016, when Justin Goldman was walking in the streets of New York City. Goldman spotted American football quarterback Tom Brady meeting with the Boston Celtics manager. He took a hurried picture with his phone and added it to the “My Story” section of his private Snapchat. Not long after, one of Goldman’s followers took a copy of the image and uploaded it to Reddit. The picture quickly began to spread across image hosting sites such as Imgur and Twitter. Soon, several news sites had published articles embedding copies of Goldman’s photo. In response, Goldman filed for direct copyright infringement based on the for-profit display of an unlicensed copy of his photograph.

What is Embedding?

Embedding is a process by which a website presents information to users without actually storing that data on its own servers. This is achieved by coding an instruction on the website that tells a user’s device where to locate the media in question. The user’s device then pulls that material through by directly accessing the servers of Facebook, Twitter or whoever else is hosting the data.

The “Server Test”

For years, media companies in the USA had defended embedding using a “Server Test”, based on a 2007 ruling from the Ninth US Circuit Court of Appeals in Perfect 10 v. Amazon (508 F.3d 1146). There, Google and other defendants were found only to be liable for direct copyright infringement in respect of images that were stored on their own servers. Companies would not be liable for images that they had merely linked to. In the words of the court, “[p]roviding HTML instructions” was not “equivalent to showing a copy.”

However, in her recent decision in Goldman v Breitbart, US District Judge Forrest was “skeptical that Perfect 10 correctly interprets the display right of the Copyright Act”. She claimed that even a “quick survey reveals that the case law in this area is somewhat scattered”. In Judge Forrest’s view, “a website’s servers need not actually store a copy of the work in order to ‘display’ it”. On the basis of the “plain language” and “legislative history” of the US Copyright Act, she concluded that physical location or possession of an image did not determine who legally ‘displays’ it.

Even if the Server Test was valid, she considered it would not have been applicable on the facts. Judge Forrest emphasised the media outlets’ conscious decision to embed the copyrighted image in their Tweets. This was in contrast to the passive style of linking used by Google.

An Intentional Challenge to Direct Infringement

The traditional Server Test specifically leaves open the possibility of establishing secondary liability in cases of linking. Unlike direct liability, which only applies to those actually doing the unauthorised copying or distributing, secondary liability extends to anybody that knowingly assists another in performing an unauthorised act.

Secondary liability would certainly have been an easier argument for Goldman to run. The news sites were providing instructions telling users how to access infringing media. The case would then centre on defences such as fair use. However, the decision to only allege direct infringement shows clear intent from Goldman’s lawyers to challenge the Server Test.

Conflicting Case Law

Judge Forrest issued permission to appeal her judgment on 19 March 2018. The defendants, who include Boston Globe, Gannett, Time Inc, Vox and Yahoo, argue that the ruling will “cause a tremendous chilling effect on the core functionality of the web”.

One such organisation, BoingBoing.net claim that the decision “rests on a serious misapprehension of the process of embedding”. They reason that linking simply makes users aware that content exists. Direct liability should not apply as only the host is actually copying or distributing anything.

BoingBoing.net themselves were involved in a similar suit that had been thrown out by a Californian District Judge a week prior to the Goldman case. The action for breach of copyright had been brought by Playboy after BoingBoing.net had published a link to a third party site hosting a historical collection of Playboy centrefolds. The case was dismissed on the basis that merely providing “the means to accomplish an infringing activity is insufficient to establish a claim for copyright infringement”. Liability would only exist “if the defendant engages in personal conduct that encourages or assists the infringement”. Depending on the outcome of the Goldman appeal, US internet users are faced with two seemingly irreconcilable interpretations of federal law.

Staying Up To Date

The conflicting interpretations of US copyright law reflect an area that is constantly forced to innovate to keep up with the times. Judge Forrest begins her judgment by writing that “[w]hen the Copyright Act was amended in 1976, the words “tweet,” “viral,” and “embed” invoked thoughts of a bird, a disease, and a reporter”. She acknowledges that changes in “technology and terminology” can mean that “from time to time, questions of copyright law will not be altogether clear”. In such a situation, courts must look to the “guiding principles of copyright”.

The accuracy of this statement can be traced back to the very inception of US copyright law. The first copyright statute in the US, enacted in 1790, designated only “maps, charts, and books”. Since then, the courts and legislature have had to adapt in order to facilitate new technologies. An example is the introduction of live video broadcasting. Congress faced the issue of reconciling the transmission of live footage with the requirement that works copyrighted under federal law be “fixed in a ‘tangible medium of expression’”. The problem, as identified in a 1976 House Report, was that the work of the cameraman and directors plainly “constitutes ‘authorship’” but did not fit within the traditional test. As such, a statutory definition of “fixation” was added that was capable of including live broadcasting.

Shift in the Copyright Landscape

In the Goldman case, Judge Forrest has paused proceedings until determination of the appeal. She has therefore yet to rule on a number of defences to liability that could provide an alternative justification for embedding and other forms of linking.  Amongst these defences are “a very serious and strong fair use” defence, licensing, authorisation, limitation on damages for innocent infringement and a defence for online service providers under the Digital Millennium Copyright Act. Though these may provide relief to the defendant companies, a shift from blanket protection to fact-specific defences would still drastically alter the landscape of online practices.

It would, for example, more closely align US law with the broadly equivalent EU “Right of Communication to the Public”, found within Article 3(1) of the InfoSoc Directive. Whilst the application of leading CJEU cases such as GS Media C-160/15 is by no means settled, it is clear that linking can result in a successful claim of copyright infringement in some circumstances. When determining liability, the CJEU has looked at factors such as whether the content was linked on a ‘for-profit’ basis, or whether the linking entity was aware that the material was infringing copyright. While these tests can offer substantial protection, a move towards this approach would place US internet users in murkier waters than ever before.

Professional or Amateur Infringers?

Should Judge Forrest’s ruling be upheld, US internet users who embed or link to digital media will need to apply a degree of discretion. This will be particularly detrimental to those organisations that currently choose to automate the process. As stated by Goldman’s lawyers, “[i]f you help yourself to any copyrighted material wherever you see it… you run the risk of being an infringer.”

The US appeal courts are faced with an unenviable task. On one hand, it is not unreasonable to expect professional media outlets to verify the legality of their sources. On the other, it is unrealistic to expect the average internet user to check that every image they post has been properly licensed.

Given the importance of the issues at hand, it is possible that the matter will ultimately end up before the Supreme Court. If so, the shared scholarly interest between EU and US copyright lawyers will result in any ruling having global repercussions. In any event, the legality of embedding and linking is likely to remain a live issue for years to come.

 

Links:

Goldman v Breitbart – Opinion & Order – February 15 2018

Goldman v Breitbart – Opinion & Order – March 9 2018

Perfect 10 v. Amazon

Playboy v Happy Mutants (BoingBoing.net)

Copyright Law Revision, September 3 1976 Report

InfoSoc Directive

GS Media

Share our insights

One comment

  1. I think the courts are going to continue to wrestle with this issue as technology evolves. Copyright lawyers will be familiar with the “if its worth copying, its protected” maxim, that helps with the substantial copying test. But linking has been a difficult question for a long time. Remember when Yahoo! was “Yet Another Hierarchical Officious Oracle” or a big collection of links? The difference there was that Yahoo (and other listing sites) are collecting links and then passing traffic to the “linkee”.

    What becomes difficult is when more information is pulled through as a result of the link and it becomes more than a description of the target of the link. So drawing through summaries (or images) and then selling a compilation of links becomes more problematic.

    The tension arises because the target wants the links – they want the traffic. But they don’t want the linker to use so much of the content that the browser doesn’t need to visit the source of the content.

    Am keen to hear other views.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.