2.01.2006

Copyright Clash Over Image Searches: An Imperfect Means to a Pornographic End?

(This article was originally featured in Modern Practice, Findlaw’s Law and Practice Technology Magazine, in February 2006.)

Everyone knows that “Pornography” (content) and “Google” (a means to online content) have played major roles in driving the development of the Web and Internet law. It may come as little surprise, then, that these two forces were recently at odds in pleadings, briefs, and a district court decision involving a search engine’s right to display copyright protected images (1) as thumbnails, and (2) through in-line linking to third party websites.

Perfect 10 publishes adult content and owns thousands of copyright protected adult images available online. Google, the largest Internet search engine, provides an “image search” function, which displays thumbnail versions of images cached on Google’s servers and inline links to images stored and served by third party websites. Google’s image search makes it easy to find Perfect 10’s images online.

In a 12-count complaint, Perfect 10 claimed, inter alia, that Google directly, vicariously, and contributorily infringes Perfect 10’s copyrighted images by displaying and distributing them via Google’s image search function. Google argued that its use of Perfect 10’s images is protected under the fair use doctrine. On Perfect 10’s motion for a preliminary injunction, the U.S. District Court for the Central District of California held that the issues would likely be decided as follows: (1) Google’s in-line links to infringing copies do not constitute direct infringement; (2) Google’s display of thumbnails of Perfect 10’s copyrighted images does constitute direct infringement (not a fair use); and (3) Google is not secondarily liable for its in-line links or thumbnails.

Direct Liability

To determine whether Google’s in-line links and thumbnails directly infringed Perfect 10’s copyrights, the court first had to determine whether Google’s activities constituted a public display or distribution of Perfect 10’s images. To do so, the court applied the “server test.” Under the server test, “the website on which content is stored and by which it is served directly to a user, not the website that inline links to it, is the website that ‘displays’ the content.”

Since third party websites, not Google, stored and served the Perfect 10 images to which Google provided in-line links, Google did not display such images. By creating and storing thumbnails of Perfect 10’s images on Google’s servers, however, Google did display such images. Thus, the court held that Perfect 10’s claim of direct infringement based on Google’s in-line links would likely fail, whereas its claim based on Google’s thumbnails would likely prevail.

The court also found that Google did not publicly distribute copies of Perfect 10’s images by in-line linking since Google was not involved in the transfer of any files—again, the third party websites were doing all the work. Although Google may distribute the thumbnails, the court found that Google’s distribution would likely be a fair use and that the issue was moot anyway since it already found that direct infringement was likely based on a violation of Perfect 10’s display right.

The most interesting aspects of this case, however, are found in the fair use analysis; not because the analysis is groundbreaking, but because of how the court distinguishes Kelley v. Aribasoft, 336 F.3d 811 (9th Cir. 2003) (finding fair use of thumbnail images). Analyzing fair use of a copyrighted work involves examining the following factors:

1. the purpose and character of the use;
2. the nature of the copyrighted work;
3. the amount and substantiality of the portion of the work used; and
4. the effect of the use upon the potential market for the copyrighted work.

The court found that two things set this case apart from the Kelly decision. First, Google’s use of thumbnails was far more commercial than the search engine’s use of thumbnails in Kelly. The fact that Google may increase user traffic and advertising revenue from its image search was not so significant since Google does not directly profit from use of Perfect 10’s images.

The court was persuaded, however, by Perfect 10’s evidence showing third-party websites that serve infringing content and receive and display AdSense ads from Google. Google’s AdSense program allows third party websites to carry Google-sponsored advertising and share revenue that flows from the advertising displays and click-throughs. Thus, the court believed that the connection between Google’s use of Perfect 10’s images and Google’s bottom line was much stronger here.

Second, Google’s use of thumbnails harms Perfect 10’s potential market for smaller images. After filing suit, Perfect 10 entered into a licensing agreement with another company for the sale and distribution of Perfect 10 reduced-size images for download to and use on cell phones. The court reasoned that since users of Google’s image search can download thumbnails of Perfect 10’s copyrighted images for free, they will be less likely to buy them from Perfect 10’s licensee. On these bases, the court found that Google’s thumbnail use was much more commercial and harmful than the use in Kelly. Thus, Google’s fair use defense failed.

Secondary Liability

Perfect 10 failed to introduce evidence showing that Google users directly infringe its copyright protected images. So the court only addressed Google’s secondary liability vis-a-vis direct infringement by the third party websites. In analyzing whether Google was contributorily liable, the court assumed (without deciding) that Google had actual knowledge of direct infringement (the first element) since it ultimately determined that Google did not materially contribute to direct infringement (the second element). In short, the court found that Google did not materially contribute to direct infringement because the infringing third-party websites existed before Google’s image search and would continue to exist were it shut down.

Perfect 10’s claim of vicarious liability also failed. Although Google, through its AdSense program, benefits financially (the first element) from the third party display of Perfect 10’s images, it does not have the right and ability to control infringing activity taking place on third party websites. As a result, Perfect 10 would be unlikely to succeed in proving that Google can be held secondarily liable.

Conclusion

The upshot is that Google’s “loss” at the preliminary injunction stage may not be all that significant to Web based activities for a few reasons. First, using thumbnails is not per se copyright infringement—The Ninth Circuit’s decision in Kelly still holds true. The nuance that the district court focused on here was whether the search engine enjoyed a direct financial benefit as a result of displaying thumbnails that infringe another’s work. The district court found that Google did benefit a little too directly, but the Ninth Circuit may view things differently if (when) this decision is appealed.

Second, in-line linking is still okay. A website that merely points to other information, stored and served on another party’s website, should be able to avoid direct infringement liability. Had the district court taken a different approach, every in-line link would be subject to potential liability. Third, search engines are not going to be held vicariously liable simply for facilitating the discovery of information —even if that information turns out to be infringing. There has to be an additional element such as the right and ability to control the infringing environment. Obviously, search engines cannot remove all infringing content from the Web.

In sum, the vitality of Web does not appear in jeopardy as a result of the district court’s decision, even if it holds on appeal. In addition, the development of Internet law does not appear stunted. Perhaps a new branch is growing from traditional fair use analysis, but on that branch each case will really bend on the relatedness of commercial benefit from the specific use in question.