Before there was "Google Earth" there was "Google image search" and since then, whether at Yahoo, MSN.com, or nearly any other major search engine, the ability to click "images" and search the entire World Wide Web for pictures has become such a standard feature of web browsing life that we can hardly imagine life online without it. Yet a recent California court decision, Perfect 10 v. Google, Inc., 04-9484 (C.D. Cal, Feb. 17, 2006), actually threatens this new but already fundamental feature of the World Wide Web.

When searching images on the web, one issue of copyright law comes immediately to our minds: can we copy these images for our own use, for example in preparing our own PowerPoint presentations or even for posting on our web sites? The ease with which we can "right click" to "save picture as" in our browsers makes infringing the copyright in these images very tempting indeed.

But there is another copyright issue that is not so obvious, and that is, "can the search engine post these "thumbnails" in our search results without violating the image owner's copyrights? After all, copyright is supposed to protect an image from being copied, so how is it that these search engines can make small and compressed copies of someone else's images without getting permission?

Existing Law of "Thumbnails"

Current law, at least in the Ninth Circuit was developed in Kelly v. Arriba Soft Corporation, 336 F.3d 811 (9th Cir. 2003). In the Kelly case, the Ninth Circuit held that search results displaying thumbnail images was allowable fair use under copyright law and therefore not infringing so long as the thumbnail's hyperlink was directed toward the web site where the full size image was posted.

After the Ninth Circuit holding in Kelly, Google launched its image search engine in 2004. On November 19, 2004, Perfect 10, a company that provides adult oriented images on the Internet, filed suit against Google, claiming Google's image search violated Perfect 10's copyrights.

Recent "Perfect 10" Decision

On February 21, 2006, the California trial court judge, in Perfect 10 v. Google, Inc., 04-9484 (C.D. Cal, Feb. 17, 2006), issued a preliminary ruling that left the law of thumbnail usage somewhat in doubt until the ruling is reconsidered on appeal. According to the judge's preliminary ruling, Google, Inc.'s thumbnail image search engine likely violated copyright owned by Perfect 10.

In deciding whether Google infringed Perfect 10's copyright by displaying thumbnail images via its image search, the Court looked to the same fair use factors set forth in Kelly. According to Section 107 of the Copyright Act, the reproduction of a copyrighted work will not constitute infringement of the copyright if the use is for "criticism, comment, news reporting, teaching, scholarship, and research." This fair use doctrine also lists four factors to evaluate whether a particular use is fair: "(1) the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes; (2) the nature of the copyrighted work; (3) the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and (4) the effect of the use upon the potential market for or value of the copyrighted work." 17 U.S.C. §107.

The trial court in Perfect 10 v. Google, Inc. distinguished the facts of its case from the facts behind the holding in Kelly by carefully evaluating each of the four statutory fair use factors. According to the preliminary ruling in Perfect 10, the crucial difference between Google's use of Perfect 10's images and Arriba Soft's use of Kelly's images was that since early 2005, Perfect 10 licensed reduced-size images of its photographs to Fonestarz Media Ltd. of the U.K. Fonestarz markets reduced-size copyright images for download and use on cell phones worldwide. As part of this licensing agreement, Fonestarz sells on average 6,000 images per month in the U.K. alone. The same reduced-size images in which Perfect 10 holds copyright registrations were also available for downloading from Google’s image search engine. The only difference between obtaining the images from Google versus Fonestarz is that the images were free from Google and one must pay to obtain the images from Fonestarz.

Analysis of Fair Use Factors as to Google’s Thumbnail Images

The first fair use factor is the purpose and character of the use. The trial court in Perfect 10 evaluated two separate aspects of this fair use factor: (a) commercial versus noncommercial use; and (b) transformative versus consumptive use. Google’s use was held to be commercial due to the fact that Google derives commercial benefit every time a user clicks on a website of a company that is part of its AdSense program. Therefore, if showing a thumbnail causes a user to click on the full-size image and then to the relevant website, and that company participates in Google’s AdSense program, Google splits the advertising revenue made from that user’s traffic. Secondly, Google’s use of the thumbnail images, although transformative, was also found to be consumptive use because the thumbnails displayed by Google can be downloaded from Google image search onto a cell phone in the exact same manner as those images licensed and sold by Fonestarz. Therefore, the Court concluded that the first factor weighed in favor of Perfect 10.

As for the second factor, the nature of the copyrighted work, the trial court found that the fact that Perfect 10’s photographs “are of scantily-clothed or nude women is of no consequence; such images have been popular subjects for artists since before the time of ‘Venus de Milo.’ ” The trial court instead looked at whether Perfect 10’s works were published or unpublished. Since Perfect 10’s images have previously been published throughout the internet and in print, the first appearance of Perfect 10’s expression of art has already occurred. The trial court then concluded that this second factor weighed in favor of Perfect 10.

The amount and substantiality of the portion of the work used is the third fair use factor. The trial court held that, just like in the Kelly case, Google’s use of Perfect 10’s images in their entirety in its image search was necessary to achieve effective image search results. The Court also stated “photographs of nude women can, like photographs of the American West [in Kelly], vary greatly.” Therefore, Google’s use of the complete images was necessary. Thus, the Court decided that the third factor did not favor either Google or Perfect 10.

Lastly, the Court evaluated the effects of Google’s use on the potential market for and value of Perfect 10’s works. This factor, along with the first and second factors, also weighed in favor of Perfect 10 because Google’s thumbnails could be downloaded to a user’s cell phone for free instead of paying for the thumbnails from Fonestarz.

Therefore, although the trial court was following the Ninth Circuit’s statement of the applicable legal principles and factors, in the Perfect 10 case the facts were different, allowing a different result. Since three of the four fair use factors weighed in favor of Perfect 10, the Court found that Google’s use of the reduced-size thumbnails was not fair use of Perfect 10’s copyrighted works.

Thumbnails Now But Not Forever?

The Perfect 10 decision in February was only preliminary, and it has been appealed by Google, so this particular decision could conceivably be overturned by the Ninth Circuit. But there is a very significant perspective that should have been addressed in both the Kelly and Perfect 10 decisions and which apparently was not.

The fact not addressed by either court is fundamentally critical to the very nature of the search engine, what it does and how we rely upon it. Court cases are decided by judges based on the facts presented by the two parties to the lawsuit. But search engines search the entire web for their results, and therefore, in any one search by Google for images, large numbers of images, all from different web sites, are posted. As anyone who has ever used an image search knows immediately, the thumbnails posted as the results do not just show the images of a single future plaintiff, who may or may not be selling the images, but rather the search results show scores, sometimes hundreds or even thousands, of images from almost as many different web sites.

The very nature of the search engine, therefore, makes the simple weighing of factors between a single image owner and the search engine the wrong path to follow. The legal result of applying four fair use factors on a case-by-case basis would result in a completely unworkable legal framework. For example, in any single image search, if one follows the reasoning of the trial court, Google could have had factors weighing in its favor vis-à-vis hundreds of potential plaintiffs and weighing against Google in hundreds more. There literally would be no possible image searching on the internet that could ever be free of risk because there is no way to limit the results to just web sites where the four fair use factors weigh in favor of Google. Under this case-by-case factor analysis, there is no current method of compliance that would enable image searching.

By way of example, publishing a poster of the Chrysler Building in New York City might or might not be fair use of the copyright in the design and image of the Chrysler Building, depending on the weighing of the fair use factors between the copyright owner and the defendant. However, with the case of Google, the overall search results of images from thousands of web sites is more like a poster of the entire New York skyline, of which images the view of the Chrysler building only forms a very small part. Applying a building-by-building test of fair use factors would undoubtedly produce different results, as some buildings may have no copyrightable elements to protect, while others are only partially shown, peeking over and behind others. The consequence of such a building-by-building fair use test would produce only one certain result—no more skyline images of New York. Is that the right result?

What the Ninth Circuit needs to do, in our opinion, is recognize the inadequacy of the factor analysis for evaluating search engine results on a case-by-case basis with a single plaintiff and a single search engine. We think the Ninth Circuit should adopt a more macro method of evaluating the four statutory fair use factors by considering the “particular use” as not being the thumbnail posting of images from a single web site, but the aggregate search results, showing thumbnails of hundreds of different web sites. In this way, to use the analogy above, we are not evaluating fair use building-by-building, or image-by-image, we are rather weighing the factors of the entire search results, like the skyline of New York City.

It is still possible that after weighing the four factors with respect to all the web sites whose images appear in the search results and giving due weight to the “necessary” method of displaying the images by the image search engine, the case could still go against Google. We don’t think it would, but it might. But such a decision, truly weighing the overall circumstances of internet web search methods and results, even if in a ruling against Google, might provide a legal rule or standard that could be complied with, without threatening the research value to us all of image search tools.