Mark Punzalan & Stephen Stanwood. Mark Punzalan is the founder of Punzalan Law, P.C., an Intellectual Property and Commercial Litigation boutique firm in Redwood City, California. Stephen Stanwood is a former Legal Extern at Punzalan Law and is the Editor-in-Chief of the Santa Clara High Tech Law Journal and Tech Law Questions.
Over the past two decades, the search engine has revolutionized the way that billions of people access information. Today, the classic search engine like Google is just the tip of the iceberg. Internet users now access third-party content online through countless intermediary websites on a daily basis. For instance, Facebook feeds are filled with snippets of articles originally published elsewhere. Similarly, Google News takes content from traditional media outlets like CNN, The Washington Post, and The New York Times and amalgamates it all onto one easy-to-read screen. Flipboard, a more recent news aggregator, takes that same concept a step further, allowing users to design their own attractive online news magazine containing articles of their choosing from sites around the Internet.
To someone unfamiliar with the current state of U.S. copyright law, much of this resembles copyright infringement. Traditionally, once a newspaper like The New York Times published an article, that paper retained the control to reproduce and distribute copies of the article. The Internet has always made life difficult for traditional copyright law, though, because some believe the Internet exists largely to distribute copies of existing materials. Copyright laws provide some guidance for online service providers to ease this natural tension between the traditional definition of copyright infringement and new content-sharing technologies.
Website operators looking to offer services relying on third-party content need to be aware of both the risks and legal protections available under copyright laws. This article provides an overview of the protections available to website operators and concludes with a discussion of the current state of the line between fair use (which is legal) and copyright infringement (which is not).
CDA § 230 Does Not Protect Against Copyright Infringement
It is first important to note one protection that is not available to website operators: § 230 of the 1996 Communications Decency Act. Generally speaking, this statute offers rather broad immunity to service providers arising from third-party information. Commentators have gone so far as to call CDA § 230 the law that “cleared the way for the modern Internet,” as the immunity it provides is a huge asset to sites like Google and Facebook that reproduce large amounts of material from third parties. Unfortunately for online service providers concerned specifically about copyright infringement liability, CDA § 230 expressly states that it has no effect on “any law pertaining to intellectual property.” Copyright law is certainly intellectual property law, so this provision does nothing to help shield providers from potential infringement liability related to third party content.
DMCA § 512 Offers Some Protection for Copies Made by an “Automatic Technical Process,” But Websites Must Cooperate with Takedown Notices
In 1998, Congress passed the Digital Millennium Copyright Act (DMCA). Though some DMCA provisions broaden definitions and increase damages for copyright infringement, the DMCA also offers “safe harbors” to certain types of online service providers. Specifically, DMCA § 512 insulates any site that uses an “automatic technical process” to make “intermediate and transient storage” of copyrighted materials from infringement liability. This can be a valuable tool for web providers who make temporary copies of third party materials. In 2006, for instance, a federal court held that Google’s 14-day caching of webpages qualified as “intermediate and temporary storage” sufficient for Google to avoid liability for copyright infringement.
More recently, in Avdeef v. Google, Inc., Google unconventionally relied on another DMCA safe harbor provision to escape infringement liability. To impose liability under § 512(c)(1), a plaintiff must show that a service providers has “actual knowledge” of infringing content and essentially did nothing in response to a takedown request. In this opinion, the plaintiff had used self-publishing service LuLu to sell his novel “The Last Breath of Mars” online and alleged that the entire eBook later appeared on Google Books for free. The court found that Google had no actual knowledge of the alleged infringement and had acted pursuant to its licensing agreement with LuLu, so it granted summary judgment for the search engine.
Avdeef reveals an important limit of DMCA § 512: once a website operator receives notice that it has posted copyrighted material (and therefore has “actual knowledge” of it), the website must take down that content. Even though Google did not incur infringement liability for originally posting Avdeef’s book, it still had to take the book offline after receiving his notice.
Fair Use is the Broadest Protection Available to Online Service Providers, But Its Exact Scope Can Be Difficult to Predict
As part of the 1976 Copyright Act, Congress codified “fair use” as a limit on the power of rightsholders to control their works. United States courts had long recognized exceptions to copyright infringement such as authors quoting limited portions of other works to review them. Congress created a test for courts to use to draw the line between “fair use” and copyright infringement. The key legal question in this test is whether the alleged infringer uses the copyrighted material in such a way that is sufficiently “transformative” for a court to call the service it provides non-infringing “fair use” rather than copyright infringement.
Traditional search engines like Google have been very successful in persuading courts to find their use of third party content as sufficiently “transformative.” Google, for instance, need not fear a copyright infringement lawsuit from the owner of every copyrighted image that appears in a Google Images search. In Kelly v. Arriba Soft, the Ninth Circuit held that the now-defunct search engine ditto.com could display thumbnail image results without incurring copyright infringement liability. More recently, in Perfect 10, Inc. v. Amazon.com, Inc., the Ninth Circuit reaffirmed this holding, specifically writing that Google’s “display of thumbnail images of photographs [constituted] fair use.”
The highest-profile search engine fair use cases are those related to the Google Books project. The search engine started Google Books in 2004 and has since scanned over 20 million physical books cover-to-cover. Google has won several important judgments against rightsholders along the way, most recently at the Second Circuit in 2015. But these decisions may not provide the clearest guidance to a new website operator looking to launch its own product.
The Recent Meltwater and TVEyes Fair Use Cases in New York Further Blur the Line Between Fair Use and Infringement for Search Engines
To add to the confusion, two recent fair use cases in the federal district court for the Southern District of New York reached opposite results on seemingly similar facts. In Associated Press v. Meltwater, the Associated Press sued a news aggregator for copyright infringement. The Meltwater website, which operated like a search engine, set up algorithms to scrape the web for current news stories, which it excerpted and sent to its subscribers. The court found Meltwater to be an infringer and that its use of news articles was not covered by the fair use doctrine.
However, the court in Fox News Network v. TVEyes went the other way. There, Fox News sued TVEyes, a company that provides a searchable database of TV and radio news, for copyright infringement. TVEyes unquestionably copied content: in fact, the company had recorded and indexed the live broadcasts of over 1400 radio and TV news stations (including Fox News) all day, every day for the past several years. TVEyes then offered that recorded content to its subscribers in both livestream and archived versions.
The court acknowledged the earlier Meltwater decision but argued that the databases in the two cases were meaningfully different. According to the court, TVEyes provided a service “that no content provider provides” while Meltwater simply “amalgamated extant content.” This is not an especially helpful distinction, as both services seem to copy similar content (news) in a similar manner. In short, these two New York decisions leave website operators may add somewhat to the confusion in the fair use / copyright infringement jurisprudence.
Website Operators Must Evaluate the Legal Risk of Using Third-Party Content and Plan Accordingly
Unless the website’s content falls squarely within the limited § 512 immunity discussed above, a web service that relies on copyrighted third party content will have to rely on fair use. But as the cases show, figuring out what exactly counts as “fair use” is a difficult exercise. From a planning perspective, companies should think very early about fair use issues. Website providers must think carefully to navigate the complex legal landscape of copyright and fair use issues.