Prof. Jane C. Ginsburg, Columbia University School of Law
January 12, 2011
The practice of online news aggregation, involving the assembly on one’s own website of information, and sometimes limited content such as headlines and lead sentences, copied from other news sources, has generated both substantial profits for the largest aggregators, such as Google News, and at least as considerable ire from some of the unsolicited sources of the copied material.1 It also has spawned several litigations in the United States alleging either or both copyright infringement and misappropriation of “hot news.” I do not propose to review the U.S. caselaw; a recent study by Kimberley Isbell of the Citizen Media Law Project 2 both describes the different types of news aggregators and examines the legal theories underlying the claims against them. Instead, I will consider the fate of similar claims recently pursued in other common law jurisdictions, Australia and the UK. The comparison merits attention not only because it illustrates the different results reached despite a common jurisprudential heritage, but also because different outcomes in different jurisdictions can complicate the business planning of web-based enterprises whose users may be located across the globe.
In Fairfax Media Pubs Pty Ltd v. Reed International Books Australia Pty, [2010] FCA 984, Judge Annabelle Bennett of the Federal Court of Australia ruled that the systematic copying of headlines from plaintiff’s financial newsletter did not infringe because the headlines, even if clever or attention-getting, were too short to qualify as protectable works of authorship. The court also expressed concern that protecting titles as works of authorship could hamper the compilation of bibliographies. The court further considered that, while the headlines were not works in themselves, neither did they combine with the text of the ensuing articles (which the defendant had abstracted) to form a single work because different groups of authors produced the headlines and the article text. While under the U.S. statute a “joint work” is formed by contributions which may be either inseparable or “interdependent,” under the Australian Copyright Act, a joint work’s contributions are “not separate from the contribution of the other author or the contributions of the other authors” (Australia, Copyright Act 1968, s 10). According to the court, the headlines were too “distinct” from the bodies of the articles for the combination to form “a discrete work in which copyright can subsist.” Thus the defendant’s service appropriated no protectable content.
By contrast, in The Newspaper Licensing Agency Ltd v. Meltwater Holding BV, [2010] EWHC 3099 (Ch), Mrs. Justice Proudman concluded that headlines and short extracts could be protected as literary works. The case concerned the activities of the contemporary version of a clipping service that copied the headlines and opening text of web-posted news sources as well as short extracts of text surrounding the search terms corresponding to the service’s clients.3 Although the defense urged that the Fairfax decision should be persuasive authority on the copyrightability of headlines, Mrs. Justice Proudman rejoined that the European Court of Justice’s 2009 decision in Infopaq International A/S v. Danske Dagblades Forening, Case C-5/08 (16 July 2009) had for the UK “overtaken” the traditional approach of Commonwealth courts. Infopaq also involved a digital clipping service, which systematically copied 11-word extracts from the plaintiff’s newspaper, consisting of the search term and the five words on either side. The ECJ ruled that the excerpts could constitute actionable reproductions “if that extract contains an element of the work which, as such, expresses the author’s own intellectual creation.” As a result, explained the English court, “the ECJ makes clear that originality rather than substantiality is the test to be applied to the part extracted.” Under that standard, at least some headlines would surely qualify; they “involve considerable skill in devising and are specifically designed to entice by informing the reader of the content of the article in an entertaining manner. [They therefore] are capable of being literary works, whether independently or as part of the articles to which they relate.” With respect to the extracts, “The effect of Infopaq is that even a very small part of the original may be protected by copyright if it demonstrates the stamp of originality reflective of the creation of the author or authors of the article. Whether it does so remains a question of fact and degree in each case.”
The English court next addressed the application of the “fair dealing” exception to the copying of headlines and extracts. “Fair dealing” in the UK copyright law bears some kinship to the U.S. “fair use” defense, but it is much more narrowly drafted. First, the defense applies only to certain activities: “research for a non-commercial purpose” and purposes of “private study” or of “criticism or review” or “reporting current events” or “cop[ying] in the course of instruction or of preparation for instruction” (Copyright Designs and Patents Act 1988 (c. 48), s 29, 30, 32(2A)). Second, the conduct (the “dealing”) must be “fair.” Third, with respect to the purposes advanced in Newspaper Licensing Agency, “criticism or review,” and “reporting current events,” the copying must be “accompanied by a sufficient acknowledgment” of the author of the copied work.
The court held that none of these conditions had been met. The purpose of the clipping service was not to criticize but “merely to enable the End User to decide whether he wants to see the content of the articles.” Similarly, “it seems tortuous to say that copying for the purpose of seeing mentions of the End Users’ clients in the news and the stories with which they have been concerned is ‘reporting current events.’” As to the second element, “fairness” implies “an overriding public interest” in the defendant’s activities: “…a dealing by a person with copyright work for his own commercial advantage – and to the actual or potential commercial disadvantage of the copyright owner – is not to be regarded as a fair dealing unless there is some overriding element of public advantage which justifies the subordination of the rights of the copyright owner.” The private business interests of the clipping service’s clients did not meet that test. Finally, the court rejected the defense that the service’s links back from the excerpted articles to their locations on the source websites constituted “sufficient acknowledgment.” The excerpts themselves did not name the author; while following the link back to the source would reveal the author’s name, unless the user decides to seek out the source, the author’s name will remain undisclosed. To adopt the defense’s “interpretation of identification of the author for the purposes of the definition of ‘sufficient acknowledgment’ renders the requirement to identify the author virtually otiose.”
Would the analysis be different if the clipping service in NLA had been a news aggregator along the lines of Google News? The Australian Federal Court in Fairfax, assessing the applicable section of the Australian Copyright Act’s version of the fair dealing defense, considered the aggregation of financial news to constitute “reporting of news” within the meaning of the statute. The court also stated that the communication of headlines and summaries was “fair” because the defendant’s service did not substitute for reading the targeted sources but rather enabled users to go to the source sites to read the stories in full. The court deemed the headlines to “come[] within the category of a citation to the source newspaper article” especially when the summaries themselves were not alleged to infringe, and the headlines served to identify the object of the summaries. The factual differences between the English and Australian cases could suggest that an aggregation service available to the public at large, and perhaps free of charge (apart from advertising) would enjoy a stronger claim in the UK to a fair dealing exception. Nonetheless, it is not clear that an English court would perceive an “overriding public interest” in “enabling the End Users to decide whether or not the articles themselves are worth reading.” The NLA court appears to discern a greater likelihood than did the Fairfax court that the aggregation of headlines and verbatim excerpts (not abstracts, as in Fairfax) would substitute for consultation of the source articles (and thus deprive the source sites’ advertisers of exposure to potential consumers).
For now, the English and Australian authorities differ on whether the aggregation of headlines and excerpts reproduces copyrightable material at all, and, even assuming copyrightability, whether the copying would constitute “fair dealing.” For present purposes, we can say that, in the United States, whatever the analysis under copyright law, some jurisdictions may consider the systematic copying and communication of time-sensitive news to be actionable misappropriation. (In the UK and Australia it appears that concepts of unfair competition are primarily confined to instances of “passing off,” and would not offer an alternative theory, along the lines of the U.S. decisions in International News Service v. Associated Press, 248 U.S. 215 (1918), and progeny,4 to combat copying.) What does this mean for a web-based news aggregator? Does it suggest that the aggregator should move its headquarters and servers to Australia to avoid liability? Whatever the attractions of Down Under as a copyright haven (or in general), relocating one’s business to those apparently friendlier climes is unlikely to absolve the aggregator if its users are accessing the service from jurisdictions in which the provision of headlines and extracts would be considered infringing. The likelihood that the aggregator would find itself held to the copyright standards of another jurisdiction (as well, perhaps, as its unfair competition norms) – at least with respect to the making available of its service to users located in that jurisdiction – is even greater if the works from which the aggregator draws were also created or published by authors and businesses located in that jurisdiction. (For further discussion of the choice-of-law issues that arise in cases alleging Internet-based infringement, see my earlier post, Conflict of Laws in the Google Book Search: A View from Abroad.)