A New York district court recently held that a host of online news publishers and media websites that embedded certain tweets (containing unauthorized uploads of plaintiff’s copyrighted photo) on their websites violated the plaintiff’s exclusive display right, despite the fact that the image at issue was hosted on a server owned and operated by an unrelated third party (i.e., Twitter). (See Goldman v. Breitbart News Network, LLC, No. 17-03144 (S.D.N.Y. Feb. 15, 2018)). In doing so, the court declined to adopt the Ninth Circuit’s so-called “server test” first espoused in the 2007 Perfect 10 decision, which held that the infringement of the public display right in a photographic image depends, in part, on where the image was hosted.
Under the “server test,” only a server that actually stored the photographs and “serves that electronic information directly to the user (`i.e., physically sending ones and zeroes over the Internet to the user’s browser’) could infringe the copyright holder’s rights.” In its ruling, the Goldman court granted the plaintiff’s motion for partial summary judgment, and determined that the reasoning of the Perfect 10 decision, which applied to a search engine’s image search function and display of thumbnails to a user, was not applicable to the embedding practices the media sites engaged in.
This decision may be an awakening of sorts for web publishers who have been under the impression that the server test was established law throughout the country or that they enjoyed near-free reign to inline link to user posts and display outside content on their own websites without fear of liability, as long as they used the source platform’s embed tool and didn’t host such content on their own servers. Embedding or inline linking to outside content is a common practice for websites or blogs (some popular sites have built a business model solely based on such practices), and the ultimate ruling in this case will be closely-watched and may end up in the Second Circuit. If the ruling stands, it could greatly affect online publishing and the free flow of content, as well as how web publishers perform due diligence before displaying social media or user-generated content containing photos or videos.
In Goldman, a photographer posted a candid photo of a group of well-known professional athletes to Snapchat. Feeding into a news story about how a football star was perhaps attempting to recruit a free agent basketball player for his hometown team, the plaintiff’s copyrighted photo went viral, moving from Snapchat to Reddit to Twitter, where it was uploaded and tweeted out by several users without plaintiff’s permission. Eventually, the tweets containing the image made their way onto the websites of various news publishers and media websites, which wrote about the developing sports story and embedded the tweets containing the photo into the articles. None of the defendant websites copied and saved the photo onto their own servers, rather making the photo visible via Twitter’s embed tool.
The plaintiff, claiming he never publicly licensed his photograph or posted it to Twitter, filed suit against the defendant websites, claiming a violation of his exclusive display right under the Copyright Act. The defendants argued, among other things, that, under the server test, they could not be held liable for infringement for images stored on a third-party server. The parties agreed to divide the litigation into two phases, with the court initially to decide the display right issue and the server test defense, and later to decide on the remaining issues.
The defendants urged the court to apply the server test to limit the scope of the display right, contending that the Ninth Circuit’s ruling was “well settled” law and merely providing HTML instructions does not constitute a “display” of the image. The court disagreed, stating that the Ninth Circuit’s decision in Perfect 10 involved connections between three parties — end users, third-party servers hosting copyrighted material, and search engines that did not themselves host copyrighted material but that linked end users to the third-party servers that did host copyrighted material — and the “server test” was a test to distinguish between the hosting third-party servers and the non-hosting search engines. The court also demonstrated through reviews of cases outside the Ninth Circuit applying the server test to the display right that this area of the law was not really “well settled.”
In sum, the court declined defendants’ invitation to apply the server test for two reasons. First, the court saw nothing in either the text or legislative history of the Copyright Act suggesting that physical possession of an image is a necessary element to its display for purposes of the Act. (Moreover, in looking to Supreme Court precedent, the court read Aereo, while not directly on point, as strongly supporting plaintiff’s argument that “liability should not hinge on invisible, technical processes imperceptible to the viewer.”) Second, the court stated that there are critical factual distinctions between Perfect 10 and the instant case such that, even if the Second Circuit were to find the server test consistent with the Copyright Act, it would be inapplicable here. In this court’s view, while the Ninth Circuit did not cabin its adoption of the server test to search engines, it relied heavily on that context in its analysis.
“[E]ven if it correctly interprets the Act, to the degree that defendants interpret Perfect 10 as standing for a broadly-construed Server Test, focusing on the physical location of allegedly infringing images, this Court disagrees. Rather, Perfect 10 was heavily informed by two factors—the fact that the defendant operated a search engine, and the fact that the user made an active choice to click on an image before it was displayed—that suggest that such a broad reading is neither appropriate nor desirable.”
Even though the court granted the plaintiff partial summary judgment on the display right issue, the litigation is far from over. The court even noted that the defendants have “a number as of yet unresolved strong defenses to liability” apart from the server test issue and even opined that “there are genuine questions about whether plaintiff effectively released his image into the public domain when he posted it to his Snapchat account.”
As discussed above, if the lower court’s ruling stands, it may affect the freewheeling use of embedded content and compel some web publishers to change diligence procedures before using embedded images, or even choose to decline to display images when they embed social media posts or similar content. For example, by default, photos, videos or links that are part of a tweet are displayed in embedded tweets, but website developers can hide such media “if editorially desired.” It is also possible that some web publishers will maintain their current processes with respect to such content, relying in part on their adoption, and the adoption by the source servers, of DMCA safe harbors. However, it is reasonable to expect that the decision will prompt web publishers and their counsel to reexamine policies surrounding the use of embedded images and user-generated content in general.