Publications

Greetings, Court Fans!

Just one decision to report in this Update, but it’s a biggie, at least for IP nerds and coders. In Google LLC v. Oracle America, Inc. (No. 18-956), the Court held that Google’s copying of certain portions of a computer program owned by Oracle, called Java, was fair use under the copyright laws. In doing so, Justice Breyer, writing for a 6-2 majority, also offered insight into how the Court would address the somewhat awkward fit for computer programming within copyright law.

The case concerns Google’s Android platform, which now earns it billions of dollars a year in revenue. Google acquired Android, Inc., a small startup, in 2005, hoping to develop a software platform for mobile devices like smartphones. In Google’s vision, the new Android platform would be free and open, so that software developers could use the tools found there to develop more and more Android-based applications, making Google’s smartphones more marketable. To help ensure that a critical mass of programmers would engage the platform, Google originally considered licensing the entire Java platform from Sun Microsystems (Oracle’s predecessor), because that platform was known and used by millions of programmers. But those talks broke down, and Google decided instead to build its own platform using a large team of Google engineers working for years to write millions of lines of code. Because Google wanted programmers familiar with Java to be able to easily work with this new Android platform, however, it copied roughly 11,500 lines of code from Sun’s Java SE program.

The case largely turns on the specific code Google copied. They are part of a tool called an Application Programming Interface, or API. As the Federal Circuit explained in this case, an API is a tool that “allows programmers to use prewritten code to build certain functions into their own programs, rather than write their own code to perform those functions from scratch.” It does so by dividing and organizing the universe of tasks that a computer can perform in a particular way. Programmers can then use the API to select the particular tasks that they need for their programs or applications. For each task, there is a computer code known as an “implementing code” that tells the computer how to perform the task the programmer has asked it to perform. In order to tell the computer what implementing code to employ, a programmer using Java’s API enters a “method call,” which interacts with another set of code, known as the “declaring code,” to locate and invoke the particular implementing code that is needed to carry out the task. This declaring code thus performs two functions in the Java API: It enables a set of shortcuts for programmers, because they can pick out a particular task from the API’s task library without having to know anything more than a simple command. And second, it organizes all the various tasks the Java platform can perform into a particular structure. It is this declaring code that was at the center of the parties’ dispute. For while Google wrote its own implementing code—code that constitutes the vast majority of both the Java and Android APIs—Google copied some of Java’s declaring code, so that programmers expert in the Java programming language could use the “task calling” system they were familiar with from Java to more easily develop applications for Android. This saved them the work of having to learn an entirely new system to call up tasks.

Shortly after Oracle bought Sun in 2010, it sued Google for copyright infringement. As the case came to the Court after a decade of litigation, Oracle’s case boiled down to two questions: Whether the portion of API that Google copied is copyrightable; and, if so, whether Google’s copying was fair use. During a multi-tiered trial, a jury concluded that Google had infringed Oracle’s copyright but deadlocked as to whether Google could successfully assert a fair-use defense. The trial judge, however, concluded that none of that mattered, because the declaring code was not copyrightable in the first place. After all, Google had written its own implementing code, which constituted the vast majority of the Android platform. In the trial judge’s view, the copied material was a “system or method of operation,” which is not copyrightable. But the Federal Circuit reversed, holding that both the declaring code and its organizational structure could be copyrighted. It also concluded that there were insufficient factual findings in the record on which to base a conclusion regarding Google’s fair-use defense, so it remanded for a new trial on that question. On remand, a jury concluded that Google had shown fair use, but the Federal Circuit once again reversed, holding that the trial court should not have put the fair-use question to a jury, since it was a question of law. It then held that Google’s use of the declaring code and structure was not fair use because “[t]here is nothing fair about taking a copyrighted work verbatim and using it for the same purpose and function as the original in a competing platform.” The Supreme Court granted cert, and reversed.

Writing for a majority of six (with Justice Barrett not participating because the case was argued before she was confirmed), Justice Breyer refused to decide the principal question that Google had raised: whether the declaring code was copyrightable in the first place. Assuming that it was, Justice Breyer and the majority concluded that Google’s use of the declaring code and structure was fair. Justice Breyer expounded at some length on the contours of the fair-use doctrine and its application to computer programming. In general, the scope of copyright protection shrinks when the copyrighted material serves a “utilitarian function” or when it’s “bound up with” uncopyrightable material. But Congress has also declared that copyright law governs computer programs, which are often highly functional and bound up with uncopyrightable ideas and concepts. In this light, Justice Breyer observed that “applying copyright law to computer programs is like assembling a jigsaw puzzle whose pieces do not quite fit.” Nonetheless, he concluded, Congress’s intent was clear. “Mindful of not ‘unduly burdening users of programs and the general public’ . . . copyright ‘should not grant anyone more economic power than is necessary to achieve the incentive to create.’” And courts must effectuate this intent through case-by-case fair-use determinations. “The upshot, in our view, is that fair use can play an important role in determining the lawful scope of a computer program.”

The majority agreed with the Federal Circuit that fair use is predominantly a question of law but reversed its determination that Google had failed to show fair use. Justice Breyer discussed each of the factors typically applied in fair-use cases, dwelling in particular on a few. First, much of the value of Java derives not from the code itself, but from the fact that computer programs have learned to use it. This suggested that the copied code, standing alone, wasn’t something at the “core” of copyright protection. Second, Google’s modifications were critical. For example, Java had been geared for desktops and laptops, while Android was focused on smartphones. Similarly, the two platforms carried out their tasks in fundamentally different ways; Google, after all, had created its own implementing code. These changes were “transformative,” weighing heavily in favor of fair use. Third, Justice Breyer prioritized quality over quantity. Rather than focus on the raw amount of code—11,500 lines or “enough to fill about 600 pages in an appendix,” according to the dissent—or the amount of code copied relative to the entire platform—just 0.4 percent—Justice Breyer asked whether the copied code “consists of the ‘heart’ of the original work’s creative expression.” Here, a close look at what Google actually copied convinced the majority that it left the “heart” intact. Therefore, Google’s use of the code and structure was fair use, not subject to copyright liability.

Justice Thomas, joined by Justice Alito, dissented. The dissenters focused not on the copyrighted material, but on the practical effects of Google’s conduct. Emphasizing the market effect, Justice Thomas noted that Google’s acts let it undercut Oracle’s entire business model. For example, while Oracle charged a fee for use of Java, Google didn’t for Android; it generated revenue from advertising and data collection. And largely because the two platforms used the same declaring code, Oracle’s existing (and future) licensees flocked to Android with little difficulty, costing Oracle significant lost revenue in licensing deals. More broadly, Justice Thomas criticized the majority’s fact-sensitive take on fair use. All types of code—declaring and implementing, alike—should be afforded similar protection, he maintained. Just as Congress had made its intent clear about whether copyright law should apply to computer programs, Justice Thomas argued that Congress made this clear, too, refusing to distinguish between code that “directly or indirectly” controlled a computer. The majority got it wrong by unbundling them.

That’s all for now. We’ll be back as soon as the Court is to keep you apprised of the latest developments.

Tadhg and Dave