IS KELLY SHIFTING UNDER GOOGLE’S FEET? NEW NINTH CIRCUIT IMPACT ON THE GOOGLE LIBRARY PROJECT LITIGATION

Abstract

The Google Library Project presents what many consider to be the perfect fair-use problem. The legal debate surrounding the Library Project has centered on the Ninth Circuit’s Kelly v. Arriba Soft. Yet recent case law presents new arguments for both sides of the Library Project litigation. This iBrief analyzes two Ninth Circuit district court decisions on fair use, Field v. Google, Inc. and Perfect 10 v. Google, Inc., and their impact on the Library Project litigation.

Introduction

¶          The Google Library Project has generated mountains of academic papers and presents numerous complex legal and pragmatic issues. The project involves scanning and making available online "snippets" of literary works protected by copyright.² On both sides, much of the debate focuses on Kelly v. Arriba Soft ("Kelly II"),³ a 2003 Ninth Circuit case that held the display of thumbnail images in an Internet search engine to be a fair use. However, two recent fair-use analyses by the Ninth Circuit’s district courts regarding various aspects of Google’s search engine technology may refine and alter the impact of this precedent. In Field v. Google, Inc.,⁴ the district court for the District of Nevada ruled Google’s storage and display of websites from cache memory was a fair use. In Perfect 10 v. Google, Inc.,⁵ the district court for the Central District of California held Google’s Image Search, a search engine very similar to that in Kelly II, was likely not a fair use. These recent fair-use applications include considerations beyond those of the Kelly II court that also warrant attention in the Google Library Project litigation.

¶          Part I of this iBrief will provide a brief overview of Google Book Search, including the litigation surrounding the Library Project currently pending in the Southern District of New York. Part II of this iBrief will discuss the fair-use provision of the Copyright Act and the likelihood that the Southern District of New York will follow the Ninth Circuit’s precedent in the area. Part III examines the Ninth Circuit’s Kelly II decision and its applicability to Google’s case, and also details the technology and fair-use analyses of the Field and Perfect 10 courts. Part IV asserts that the additional considerations of these district courts should apply in a fair-use analysis of the Library Project and develops the arguments and implications Field and Perfect 10 may have on the Library Project litigation.

I. Google Book Search and the Debate Thus Far

¶          On December 14, 2004, Google Inc. announced the ambitious "Google Book Search" program (previously "Google Print"), indicating the intent of the online search engine leader to scan materials from five major libraries and make the resources searchable online.⁶ The project involves two separate programs, the "Partner Program" (formerly the "Publisher Program") and the "Library Project."

¶          Under the Partner Program, with the content owner’s authorization, Google scans the full text of a book currently under copyright protection into its search database.⁷ The database then responds to user queries by providing several pages of the book surrounding the relevant inquiry, as well as links to purchase the book from established online bookstores or directly from the publisher’s website.⁸ Recently, Google announced it would also enter the book sales market, inviting U.S. and British publishers and authors to sell online access to browser-based copies of their books.⁹ The Partner Program has not produced the legal turmoil of the Library Project because it is conducted cooperatively with content owners and similar to other competitive programs.¹⁰ 

¶          The controversial Library Project entails Google scanning materials, regardless of copyright status, from the libraries of the University of California,¹¹ Harvard University, Stanford University, Oxford University, the University of Michigan, and the New York Public Library.¹² Google saves the scanned copy of the materials on its search database.¹³

¶          In response to user queries, the displayed result depends upon the copyright status of the book. The full texts of relevant public domain materials are displayed. However, if the book is still under copyright, only a few sentences surrounding the search term, what Google defines as a "snippet," is displayed.¹⁴ Google Book Search limits the amount of snippets displayed per book to each user and does not display any snippets from certain reference books, whose purpose may be obviated by public availability of even a snippet.¹⁵ Alongside search results is information for purchasing the book, as well as the nearest available public library copy.¹⁶ No advertising is displayed with the results.¹⁷ Google also provides digital versions of the works to the respective library from which the copies are scanned.¹⁸

¶          Both the Association of American Publishers ("AAP") and the Authors Guild ("AG") responded with heavy criticism and expressed doubt regarding the legality of the Library Project.¹⁹ In response, Google announced it would suspend copying until November 2005 to provide for the implementation of an "opt-out" program, through which the owner to the rights of a protected work can instruct that the book not be included in the Library Project.²⁰ Thus, as Professor Jonathan Band explains, Google provides a copyright owner with three choices with respect to any particular work:

[T]he owner can participate in the Partner Program, in which case it would share in revenue derived from the display of pages from the work in response to user queries; it can let Google scan the book under the Library Project and display snippets in response to user queries; or it can opt-out of the Library Project, in which case Google will not scan its book.²¹

¶          Other online giants are involved in similar endeavors, although in less controversial manners. Rivals Yahoo and Microsoft MSN are working with the Internet Archive on a competing book digitization project, which will only scan public domain works or works authorized by the copyright owners.²² Public domain works have been available online for years through Project Gutenberg.²³ In November of 2005, Amazon.com announced it would sell online books through a search program similar to Yahoo’s, and Random House, the world’s largest publisher of trade books, claimed it had a business plan for allowing similar online viewing.²⁴ Yet Google’s scanning of copyrighted content without permission stands alone, as these other programs only scan public domain works or follow an "opt-in" model for copyrighted works.

¶          Google suspended copying from August until November 1, 2005, to allow for content owners to exercise their opt-out privileges.²⁵ However, alleging a "plain and brazen violation of copyright law," the AG filed suit against Google for copyright infringement in September of 2005.²⁶ Five major publishing companies, members of the AAP, quickly followed suit the next month.²⁷ Both suits were filed in the District Court for the Southern District of New York.²⁸ Google responded that its Library Project was "fully consistent with both the fair use doctrine . . . and the principles underlying copyright law itself."²⁹

II. Section 107 and the Southern District of New York

¶          "From the infancy of copyright protection, some opportunity for fair use of copyrighted materials has been thought necessary to fulfill copyright's very purpose, ‘[t]o promote the Progress of Science and useful Arts . . . .’"³⁰ Usage of a copyrighted work that would otherwise infringe the copyright holder’s exclusive rights is not an infringement if it is determined to be a fair use.³¹ Section 107 of the Copyright Act dictates four non-exclusive factors to be considered in fair use analysis on a case-by-case basis. These are: (1) the purpose and character of the use (including if it is commercial in nature or a "transformative" use); (2) the nature of the copyrighted work; (3) the amount of the work used; and (4) the effects or potential effects on the market for the original work.³² While the statutory law requires all four factors be considered in fair-use determinations, the first and fourth factors have been identified as the most important³³ and have received the greatest consideration in Kelly II and its progeny.

¶          Before addressing the Ninth Circuit’s interpretations of the fair-use statute, it is worthwhile to evaluate the likelihood that the Southern District of New York would look to the Ninth Circuit’s non-binding precedent for guidance. As discussed below,³⁴ the facts and legal analysis of the Kelly II decision provide compelling analogies to the Library Project litigation.

¶          Despite these parallels, Professor Richard Epstein has noted the differing interests of East and West Coast industries and the influence this may have on their respective circuits’ viewpoints:

The East Coast is the home of authors and publishers. Content is king so that its protection becomes a powerful institutional interest. The West Coast is the home of the persons who distribute content, like Google, not those who create it. Hence there the natural bias is in favor of allowing the free flow of information that any assertion of intellectual property rights could disrupt.³⁵

¶          While Google may be up against "the publishers’ and authors’ home court advantage,"³⁶ other circuits have seen fit to follow Ninth Circuit precedent in pragmatic Internet-related copyright issues.³⁷ Kelly II has been cited in various fair-use analyses of district courts in other circuits, including in the Southern District of New York³⁸ and is prevalent in the commentary, press releases, and academic debate surrounding the Library Project.³⁹ 

III. Kelly II Precedent and New Ninth Circuit Developments

¶          While fair use is an "equitable rule of reason" that must be assessed "on a case-by-case basis,"⁴⁰ many commentators have looked to a 2003 Ninth Circuit opinion, Kelly v. Arriba Soft Corp. ("Kelly II"),⁴¹ a seminal case involving "the application of copyright law to the vast world of the [I]nternet and [I]nternet search engines."⁴² 

¶          A pair of 2006 decisions in the district courts of the Ninth Circuit, in which Google itself was the defendant, may affect Kelly II’s impact on the Google Library Project litigation. In January, the District Court for the District of Nevada held in Field v. Google Inc. ("Field")⁴³ that Google’s online storage and display of cached websites was a fair use.⁴⁴ The next month, in Perfect 10 v. Google Inc. ("Perfect 10"),⁴⁵ based on facts only slightly different from those in Kelly II, the District Court for the Central District of California found that Google’s creation and public display of thumbnails likely did infringe the plaintiff’s copyright protection and did not constitute fair use, "despite the enormous public benefit search engines such as Google provide."⁴⁶

¶          In 2003, after withdrawing its prior decision for procedural reasons,⁴⁷ the Ninth Circuit held that Arriba Soft’s Internet search engine, which generated thumbnail images in response to user searches, was a fair use.⁴⁸

¶          In Kelly II, Defendant Arriba Soft⁴⁹ operated an Internet search engine that displayed small "thumbnail" images, rather than text, in response to a user’s search terms.⁵⁰ Thumbnail images are reduced-size, lower-resolution versions of original pictures displayed on the Internet.⁵¹ Arriba downloaded the full-size original images from the webpage in order to create the thumbnails and subsequently deleted the originals after storing the thumbnails on Arriba’s server.⁵² If a user selected a particular thumbnail image indexed in the result, the search engine would display the full-sized image, surrounded by explanatory text, Arriba-supplied advertising, and a link to the original web site.⁵³ The full-sized image was incorporated into a display through a process known as "in-line linking," which displayed the image in Arriba’s page directly from the originating website without ever copying it onto Arriba’s server.⁵⁴ The court noted that "as a result . . . the user would not realize that the image actually resided on another web site."⁵⁵

¶          When Plaintiff Kelly, a photographer who displayed copyrighted images on his own website and licensed his images to others, complained about his images being indexed on Arriba’s search engine, Arriba deleted links to Kelly’s own website.⁵⁶ Kelly subsequently sued Arriba for copyright infringement stemming from images copied by the Arriba software off of licensed third-party websites.⁵⁷ Arriba moved for summary judgment based upon fair use, and the district court found both the thumbnails and in-line linking to be fair uses⁵⁸ under Section 107 of the Copyright Act.⁵⁹

¶          In Kelly II, the Ninth Circuit analyzed the four statutory fair use considerations and determined that Arriba’s use of Kelly’s images was a fair use. Regarding the purpose and character of the search engine, the Ninth Circuit held that, although Arriba operated its website for commercial purposes and its use of Kelly’s images was commercial, the use was "more incidental and less exploitative in nature than more traditional types of commercial use."⁶⁰ Despite Arriba-placed advertising appearing on the search results webpage and in-line linked pages, the court found that "Arriba was neither using Kelly’s images to directly promote its web site nor trying to profit by selling Kelly’s images," and instead, "Kelly’s images were among thousands of images in Arriba’s search engine database."⁶¹ The court also found the thumbnail images transformative and held that this mitigated the significance of the commercial purposes.⁶² In fair-use analysis, a transformative use is one that rather than "merely supersed[ing] the object of the originals. . . serves an entirely different function than [the original use, in this case] improving access to information on the Internet versus artistic expression."⁶³ The thumbnails also were not likely to be used for aesthetic purposes because of their smaller size and inferior resolution.⁶⁴

¶          Also within its discussion of the first fair use factor, the court gave weight to the public benefit bestowed by enhanced information-gathering techniques on the Internet.⁶⁵ Analogizing the search engine to comparative advertising, the court held that the functional benefit of Arriba’s use compared with the "minimal loss of integrity to Kelly’s images" further inclined the first Section 107 factor towards fair use.⁶⁶

¶          Under the second statutory factor, the Ninth Circuit determined that Kelly’s works were clearly creative in nature. However, this weighed only slightly against fair use, as the works had already been published broadly on the Internet by Kelly.⁶⁷

¶          The third factor was neutral.⁶⁸ Considering the amount of the copyrighted work that Arriba used, the court found that, although Arriba copied Kelly’s works in their entirety, this wholesale copying was reasonable in light of Arriba’s intended use.⁶⁹ The court noted that if "Arriba only copied part of the image . . . [it would reduce] the usefulness of the visual search engine."⁷⁰

¶          Finally, under the fourth Section 107 factor, the Ninth Circuit held that Arriba’s creation and use of the thumbnails did not harm the commercial value of Kelly’s images.⁷¹ The court found that Arriba’s thumbnail results "would guide users to Kelly’s web site rather than away from it," thus not detracting from selling or advertising on the original website.⁷² While making clear that the inferior quality of a reproduction may not always assist an alleged infringer in demonstrating fair use,⁷³ the court noted that if a user wanted to view or download a quality image, he would have to visit Kelly’s website.⁷⁴ Further, at the time, Arriba did not sell or license its thumbnails to other parties, and no foreseeable market for thumbnail-versions of the images existed.⁷⁵

¶          After weighing the four Section 107 fair use factors, the Ninth Circuit in Kelly II found that Arriba’s generation and display of thumbnail images was a fair use under the Copyright Act, despite its incidental commercial purpose and wholesale copying of Kelly’s creative works.⁷⁶ 

¶          In Field, the District Court of Nevada held that Google’s cache storage did not create liability for direct copyright infringement because the display of cached versions of web pages was fair use.⁷⁷

¶          Google’s main search engine⁷⁸ scans the web using an automated search tool commonly known as a "web crawler."⁷⁹ The web crawler continuously scans available pages on the Internet and catalogs those pages into Google’s searchable web index.⁸⁰ As a part of this process, Google also stores the website’s source code in a temporary repository on its servers, known as a "cache."⁸¹ When Google’s search engine runs a text search, links to the cached copy of the websites are returned in the index alongside the full URL for the original page. By clicking on the cached link rather than the original URL, the user views an archival copy of the webpage, "as it appeared the last time the site was visited and analyzed by the [web crawler]."⁸² When a cached page is displayed, a large disclaimer explains that the user is not at the original website and two hyperlinks lead to the original website. Other major search engines, including Yahoo! and MSN, also feature cached links. Although Google has provided its cache since 1998, this was the first suit stemming from their inclusion.⁸³

¶          As the court noted, Google’s web crawling is an automated process.⁸⁴ Inevitably, in scouring the Internet, the search engine will analyze copyrighted content placed on websites, and through the cache process, save a copy of this content to Google’s servers. The Nevada District Court stated "[g]iven the breadth of the Internet, it is not possible for Google (or other search engines) to personally contact every Web site owner to determine whether the owner wants the pages in its site listed in search results or accessible through ‘Cached’ links."⁸⁵ To allow content owners to control both whether their websites are analyzed by the search engine and whether an archived cache copy is stored, the Internet industry "developed a set of widely recognized and well-publicized industry standard protocols by which Web site owners can automatically communicate their preferences to search engines such as Google."⁸⁶ Through a relatively easy and short process ("a matter of seconds") a website manager can place specific instruction in either the website’s meta-tags, or in a "robot.txt" file, which give instructions to either prevent archival cached copies from being generated, or prevent the search engine from indexing the website at all.⁸⁷ Further, a website owner can also make requests directly to Google, following step-by-step instructions through an online process.⁸⁸

¶          Plaintiff Blake Field, a recent law school graduate, posted fifty-one copyrighted works on his website freely accessible to the public.⁸⁹ Field "created a robots.txt file for his site, and set the permissions . . . to allow all robots to visit and index all of the pages on the site."⁹⁰ Further, although Field was aware of the "no-archive" meta-tag process, he consciously chose not to use the "no-archive" meta-tag on his website, presumably to allow the Google search engine to create cached copies.⁹¹ The Google web-crawler did just this, and once the cached links had been selected by other users (technically constituting a "display" for copyright purposes), Field filed a complaint for direct copyright infringement,⁹² seeking statutory damages in the amount of $50,000 for each of the fifty-one copyrighted works.⁹³

¶          Google moved for summary judgment based, among other defenses, on fair use.⁹⁴ The district court granted all motions in favor of Google and dismissed Blake’s case on summary judgment.⁹⁵

¶          Applying the fair-use analysis established by the four Section 107 factors, the Nevada District Court first looked to the purpose and character of Google’s use and found that, while Google did have a commercial purpose, its use was transformative in nature, thus weighing heavily in favor of fair use.⁹⁶ Regarding Google’s commercial nature, the court noted:

While Google is a for-profit corporation, there is no evidence Google profited in any way by the use of any of Field’s works. Rather, Field’s works were among billions of works in Google’s database. . . . Moreover, when a user accesses a page via Google’s "Cached" links, Google displays no advertising to the user, and does not otherwise offer a commercial transaction to the user.⁹⁷

¶          Assessing Google’s cache’s transformative nature, the court first focused on several "socially important" uses of the cache for the public benefit, including allowing access when the original page is inaccessible, allowing users to detect changes made to the particular web page over time, and allowing users to understand why a page was responsive to their original query.⁹⁸ The court also found that, by including links to the original page, the cached version would not substitute for the original, and "any user seeking to access the original page has more than ample opportunity to do so."⁹⁹ Finally, Google’s efforts to ensure that any site owner could easily prevent cached copies from being made, and the continued permission for cached links by sophisticated website owners, further evidenced that Google’s cache was not viewed as a substitute to the original web pages.¹⁰⁰ These "different and socially important purposes" minimized Google’s commercial nature and weighed heavily in favor of fair use.¹⁰¹

¶          Similar to Kelly II, the Nevada District Court placed little emphasis on the second and third factors. Although Field’s works were creative in nature, he had made them widely available on the Internet.¹⁰² While the content was copied in its entirety through Google’s cache, it was reasonable to do so, because Google’s socially valuable purposes "could not be effectively accomplished by using only portions of the Web pages."¹⁰³

¶          Finally, the court found no evidence of any market for Field’s works.¹⁰⁴ Field’s "circular" claim that Google’s use deprived him of potential revenue he could have obtained by licensing Google the right to provide cached links for his works was rejected as a "syllogism."¹⁰⁵ Further, the court found no evidence that a market was likely to develop for licensing search engines the right to allow access to web pages through cached links.¹⁰⁶ This factor thus weighed heavily in favor of fair use.¹⁰⁷

¶          While the district court’s analysis under the four statutory factors seemed sufficiently conclusive for a finding of fair use, the court proceeded to consider an additional factor, Google’s good faith, both in operating its search engine generally and with Field’s works in particular.¹⁰⁸ Google’s steps to place control in the hands of the content owners by providing several simple methods for preventing cached copies from being created, as well as its display of disclaimers and links to the original web pages, showed good faith in operating its search engine’s cache procedure.¹⁰⁹ Further, when notified of Field’s lawsuit, Google promptly removed cached links to Field’s pages without being asked.¹¹⁰ Finally, the court juxtaposed Google’s good faith with Field’s conduct and noted that the two stood "in marked contrast"¹¹¹ because Field "decided to manufacture a claim for copyright infringement . . . in the hopes of making money from Google’s standard practice."¹¹²

¶          In Perfect 10, the Central District of California held that, despite its similarity to the search engine in Kelly II, Google’s display of thumbnail versions of copyrighted images likely did not qualify as fair use.¹¹³ 

¶          Google Image Search¹¹⁴ is an online search engine that operates, at a general level, very similarly to the Arriba Soft system in Kelly II.¹¹⁵ In response to text queries, the search engine generates and returns an index of relevant thumbnail images.¹¹⁶ To generate these indexes, the thumbnails are stored in Google’s cache memory.¹¹⁷ The same methods of preventing web crawler access that apply in Google’s text-based search engine¹¹⁸ are also available for the Image Search.¹¹⁹ When the user selects a thumbnail within the results index, the original website, featuring the original image in its full-size, is incorporated into a web browser window through in-line linking a process Google calls "framing."¹²⁰ The full-size image is not stored in the cache, only the in-line link or other HTML code.¹²¹

¶          Google derives significant revenue from its AdSense advertising program. AdSense "allows pages on third party sites to carry Google-sponsored advertising and share with Google the revenue that flows from the advertising displays and click-throughs."¹²² Website owners place code on their sites asking Google’s server to select relevant advertisements to place on the webpages based on the sites’ content, and revenue is divided between the website owners and Google.¹²³

¶          Perfect 10--a content provider of "high-quality, nude photographs of ‘natural’ models," operating both an adult magazine and a subscription website¹²⁴--generates revenue from its copyrighted works in three formats; magazines, website subscriptions, and licensing agreements with Fonestarz Media Limited for reduced-size images for download and use on cell phones.¹²⁵ Perfect 10 entered into its licensing agreement with Fonestarz in early 2005, after commencing litigation against Google.¹²⁶

¶          Many of Perfect 10’s copyrighted images have been displayed without authorization on other third-party websites.¹²⁷ These third-party websites are indexed when a relevant Google Image Search is run, and thumbnails are generated and stored on Google’s servers.¹²⁸ 

¶          Perfect 10 brought suit against Google for preliminary injunctive relief, claiming direct and contributory infringement by Google for storing and displaying thumbnail images.¹²⁹ Most important to the Google Library litigation, the court held in favor of Perfect 10’s claim that Google’s display of the thumbnail images was infringing and likely not a fair use.

¶          The court first found that Google’s use was more commercial than that in Kelly II. The commercial benefit garnered from drawing increased web traffic would ordinarily fall within Kelly II’s category of "more incidental and less exploitative in nature than more traditional types of commercial use."¹³⁰ However, unlike Arriba, Google Image Search offers and derives a commercial benefit from the AdSense program. By directing users, through thumbnail indexes, to third-party websites that carry infringing copies of Perfect 10’s copyrighted images, "Google’s thumbnails lead users to sites that directly benefit Google’s bottom line."¹³¹

¶          The court also found Google’s thumbnail images to be both transformative and consumptive under the first fair use factor.¹³² The thumbnails served different purposes from, and did not supersede, the original full-sized images, qualifying as transformative under Kelly II.¹³³ Yet the thumbnail images were consumptive as to the reduced-size images that Perfect 10 licensed to Fonestarz.¹³⁴ The thumbnails were "essentially the same size and of the same quality" and could supersede the images licensed to Fonestarz.¹³⁵

¶          The second and third factors were analyzed similarly to Kelly II, leading to the same results, with these factors having little influence on the fair use determination.¹³⁶

¶          The fourth Section 107 factor also weighed slightly in favor of Perfect 10. While Google’s thumbnail images would not affect any market for the full-size originals, the court found the thumbnails likely would harm the potential market for downloading of Perfect 10’s images onto cell phones.¹³⁷ Although Google argued that, because the cell phone image download market was growing, its generation of thumbnails must not be having a negative impact, the court countered that the market "may have grown even faster but for the fact that mobile users . . . can download the Google thumbnails at no cost."¹³⁸

¶          Due primarily to the commercial purpose and dual transformative-consumptive nature of Google’s thumbnails, as well as potential for harm in the cell phone image-download market, the district court held Google’s use of thumbnail images likely did not constitute fair use and issued an injunction in Perfect 10’s favor.¹³⁹ The court reached this conclusion:

[D]espite the enormous public benefit that search engines such as Google provide. Although the Court is reluctant to issue a ruling that might impede the advance of Internet technology, and although it is appropriate . . . to consider the immense value to the public of such technologies, existing judicial precedents do not allow such considerations to trump a reasoned analysis of the four fair use factors.¹⁴⁰

¶          Sensitive to the potential for imposing broad liability for the socially valuable Image Search, the court ordered the parties to negotiate terms that would be "carefully tailored to balance the competing interests . . . of intellectual property rights on the one hand and those promoting access to information on the other."¹⁴¹

IV. Application of New Ninth Circuit Developments to the Google Library Project

¶          The Ninth Circuit’s Kelly II¹⁴² decision is compelling, though non-binding, precedent in the Library Project litigation. However, several points from Field¹⁴³ and Perfect 10¹⁴⁴ can be applied to the Google Library Project, both developing and limiting the Kelly II’s precedent regarding the first and fourth factors of the fair-use analysis.

¶          The facts and analysis of Kelly II lend well to the Google Library Project. Google’s use of the original works is commercial but arguably "more incidental and less exploitative in nature."¹⁴⁵ The display of snippets for locating materials relevant to search queries is different than the purpose and function of the original books, and the Library Project’s information-gathering function, and enormous potential for research purposes, clearly presents a public benefit.

¶          Although many of the books Google copies are creative, they have all been published (although not necessarily online) and thus do not impinge on the author’s right of first publication.¹⁴⁶ Google copies the works in full, yet such wholesale copying is necessary to create a functional search engine. Finally, while it is arguable whether the content-owners of library books may lose the licensing value of their works due to Google’s actions, the search-engine is not created to replace demand for full books and is designed to lead users to locations for purchasing the original works.¹⁴⁷

¶          Both the Field and Perfect 10 courts recognize the great public benefits search engines provide. The public benefit is frequently considered within the first Section 107 factor, essentially making a transformative publicly beneficial use stronger than a transformative personal use. The evidence of the Google Library Project’s public benefit is without question, particularly considering the recent legislative focus on orphan works¹⁴⁸,and the Library Project’s potential to revive printed literature. This may counteract any additional weight given to Google’s commercial purpose.

¶          However, the Perfect 10 court ruled against a finding of fair use "despite the enormous public benefit that search engines . . . provide."¹⁴⁹ The Central District of California refused to simply allow a great public benefit to be determinative, even when finding the statutory fair use factors weighed only "slightly in favor" of the plaintiff: "although it is appropriate for courts to consider the immense value to the public of such technologies, existing judicial precedents do not allow such considerations to trump a reasoned analysis of the four fair use factors."¹⁵⁰

¶          The Field court’s emphasis on Google’s widely published and relatively simple methods for preventing its search engine from indexing and caching website content, as well as honoring requests from content owners, may be viewed as an endorsement of an opt-out provision within fair-use analysis. The court found that the methods provided sufficient limitations for "site owners, and not Google, [to] control whether ‘Cached’ links will appear for their pages."¹⁵¹

¶          The Perfect 10 court did not lend much credence to this line of thought. However, while Field could easily prevent all unauthorized use of his works, Perfect 10 could not because the works were being indexed from third-party websites.¹⁵² Pragmatic concerns also may make providing an "opt-out" method influential, particularly in an information-gathering context where, as the Field court found, "it is not possible . . . to personally contact every [content] owner" to obtain permission.¹⁵³ This analysis is certainly relevant concerning orphan works, a particularly sensitive interest in copyright policy.

¶          As in Field, these considerations may fall under the transformative nature analysis of the first fair-use factor, or, alternatively, within consideration of good faith. Certainly, providing an "opt-out" method alone would not immunize a defendant from copyright infringement claims. However, volunteering a relatively simple and effective method for content owners to prevent their works from being included in a vast project may lessen the image of authors’ works being wrestled from their grasp. Additionally, Google’s allowance for content owners to transfer their works into the Member Program¹⁵⁴ could weigh against a finding of a consumptive use. The analysis could be given comparative value to the published status of a work, which neutralizes the weight of the works’ creative nature in the second fair-use factor.¹⁵⁵

¶          If Google’s "opt-out" offer, and temporary suspension of copying, is instead seen simply as a consideration of good faith, it will likely have little impact on the Southern District of New York’s analysis. In 2004, the Second Circuit marginalized the weight given in fair-use cases to the good faith of the alleged infringer.¹⁵⁶ The different treatment accorded to good faith by the Second and Ninth Circuits would likely preclude from the Google Library Project litigation such substantial consideration as was given by the Field court.

¶          Both the Field and Perfect 10 technologies share common factors with the Library Project’s commercial nature. The Field case featured a "more incidental and less exploitative" commercial purpose, similar to that in Kelly II.¹⁵⁷ The Field court noted that "no advertising was displayed to the user [when viewing the cached links],"¹⁵⁸ which is also true of the Library Project’s display of Book Search results.¹⁵⁹

¶          However, the Field court also noted that Google "does not otherwise offer a commercial transaction to the user."¹⁶⁰ While perhaps true at the outset of the Library Project, Google has recently announced its program allowing users to sign in and purchase browser-based access books, in a joint venture with publishers.¹⁶¹ Although the relationship between the Library Project and these new online sales is not yet clear, this may push the commercial purpose of Google’s books search beyond Kelly II’s "incidental" category.

¶          In Perfect 10, it is arguable whether the financial benefit of AdSense corresponded as directly with the thumbnail images generated from infringing websites as the court found.¹⁶² However, this shows that any benefit more direct than merely increased traffic can add weight to the commercial nature of the use. Google’s Book Search could likely be the most complete online book search index, due largely to the Library Project’s ability to bypass the inefficient and often futile search for content owners. The tie between Google’s draw of readers to its Book Search, in part resulting from the alleged infringing use, and direct financial benefits from the new business model offering online transactions brokered by Google, could shift the Library Project into Perfect 10’s realm of somewhere beyond "incidental and less exploitative" use.

¶          Perfect 10’s licensing agreement with Fonestarz shows that, despite a highly transformative use, even a licensing potential that develops subsequent to commencing litigation can weigh against the fair use of an established business practice and information-gathering system.¹⁶³ If the content owners in the Google Library litigation subsequently develop a market for licensing the display of "snippets," or even if the court finds likely a potential for such a market, the market harm factor would likely weigh against Google.

Conclusion

¶          The debate surrounding the Google Library Project has included extensive discussion of the Ninth Circuit’s Kelly v. Arriba Soft¹⁶⁴ fair-use analysis. Assuming the Southern District of New York does look to the Ninth Circuit’s Kelly II precedent, the Field¹⁶⁵ and Perfect 10¹⁶⁶ cases should guide the analysis further.

¶          While copyright is an "opt-in" system, the Field court emphasized Google’s affirmative efforts to allow website owners to maintain control over their works.¹⁶⁷ While this alone would not dictate a finding of fair use, it should at least contradict the Publisher’s assertion that Google’s "opt-out" procedure turns "every principle of copyright law on its ear."¹⁶⁸ Consideration of such affirmative efforts by defendants seems particularly prudent and logical in cases involving Internet-based technologies, where novel fair use determinations are necessary to prevent impeding the development of technology.

¶          Perfect 10’s findings of what may be defined as a "more than incidental" commercial nature, and likelihood of market harm for a newly developed licensing market created subsequent to commencing litigation, appear doctrinally correct. However, this shows the peril of entering into a large-scale endeavor relying upon a fair-use defense. Even with a project designed around the fair-use standards, the developments and uses for technology unforeseen at the project’s initiation can make a previously purely transformative use also consumptive and establish market harm. In the fair-use analysis of such technologies, as in the Library Project litigation, a court should give less weight to the subsequently developed markets, perhaps allowing for a finding of good faith in the development of the technology to mitigate the effect of the market harm.

¶          Perfect 10 also shows that, regardless of public sentiment and industry standard, even a use providing great benefit to the public must be evaluated on a case-by-case basis. The preliminary injunction against Google’s Image Search was extremely narrow, limited by the district court to terms negotiated between Google and Perfect 10 which would preserve the public benefit derived from the technology.¹⁶⁹ However, an injunction requiring Google to locate and request permission from content owners for every resource used in the Library Project would either present preventative search costs which would derail the project, or eliminate the benefits above-and-beyond other book search engines that the project stands poised to provide. Where, as in the Google Library Project, obtaining content owner’s permission would be preventatively costly and affirmative steps have been taken to allow content owners to maintain control of their works, the great public benefit of such a project is an appropriate consideration, and should weigh strongly in favor of a finding of fair use.

Footnotes