Surprising ways code-copying meets ‘fair use’ threshold

On Monday, in Google LLC v. Oracle America, Inc, the U.S. Supreme Court practically and pragmatically applied a novel extension of “fair use” to the traditional law of copyright — brightening the twilight zone incentives between copyright law and patent law protection in the case of software just a bit. It may be a harbinger of a need for the Copyright Act to be revised.

In a 6-2 decision, the Supreme Court’s ruling in favor of Google in how it developed its Android smartphone operating system shows how U.S. copyright applies to API (Application Programming Interface), software code that enables programs to work with each other, and that such new uses are truly transformational under copyright law. 

It’s an obvious conclusion considering the first principles of copyright protection. Otherwise, copyright holders, rather than relying on the renewing of creativity and originality that the copyright protects, could rest assured of continued success because no competitor has created something so original and superior that an inert user base would change products. Truly an incentive-reward to copyright holders, but one that seems out of the scope of the social contract between creators and the public that Congress struck when enacting the 1976 Copyright Act to incentivize the creation of new works.

Justice Stephen Breyer, the Supreme Court’s leading copyright scholar and jurist, observed that the law of copyright, when applied to computer programs, is like “assembling a jigsaw puzzle whose pieces do not quite fit.” Here, the Court made it fit. Although like a child forcing pieces of the puzzle together, the decision may cause distortions elsewhere in boundaries of copyright law.

Copyright law typically protects creative works, while patent law protects functional (useful) works. Computer software is especially problematic because the difference between code as “useful” when running a central processing unit (CPU) and expressive in other contexts is, at best, metaphysical at times — leaving law professors, lawyers and students at times screaming, “I know it when I see it” using the immortal words used by Justice Potter Stewart, echoing the lack of definitional precision of a frustrated Justice Stewart at the Court’s inability to define pornography.

The facts, in this case, are more problematic because Oracle’s API is more akin to a method of operation, which is clearly excluded by the 1976 Copyright Act from copyright protection.

Google needed to copy 11,500 lines (or .4 percent) of the API code to create a new transformative work — a software program that runs on mobile phones. In contrast, the original Oracle runs on desktop or laptop computers.

Few would deny that the Android operating system, representing 72 percent of the market, has changed the world, by any definition that is sufficient to be transformational and persuade a court in a copyright fair-use defense. Because Google only used enough code to create an interoperable Android version, the U.S. Supreme Court found fair use. 

In resolving this case, the Supreme Court went back to first principles and looked at the parody cases. In a parody case, the parodist needs to use the heart of the most creative part of the work to create a new work. “Fair use” accepts this copying of original expression because it is necessary to create a new transformative work. 

Similarly, here, Google needed the “heart” of the API to make Android. In the context of copyright law, the heart of the work is not stationary; instead, judicially it moves according to the alleged infringer’s value and needs. It is unlikely that earlier cases would have found the use of a copyrighted work to create a competing commercial copyrighted work a fair use.

Interestingly, the Court’s discussion of the effect on the market may have substantially changed the fourth factor of the fair-use defense. Usually, the effect on the market would limit itself to the usual markets that the copyright owner could exploit by producing the work or granting licenses. 

The Court analyzed the fourth factor as to whether Oracle could enter the mobile phone market, and it could not, so there was little effect on the market. The Court analyzed this factor almost as an antitrust issue in defining competitive markets — the computer/laptop operating system market vs. the mobile phone market, rather than using the more traditional approach at the very least, there was a licensing market for the copyrighted work.

Remember, in Campbell v. Acuff Rose, a unanimous court found a potential market for a non-parody rap version of the song “Pretty Woman,” a market that the copyright owners of “Pretty Woman” did not want to exploit.

Finally, and rather shockingly, was the court’s consideration of positive externalities of the alleged infringement, when taken to its logical extreme in the context of a traditional expressive work. For example, if you are not ready or able to make a movie about your book and I make the movie based on your book, and you then sell more books, that should weigh in favor of my infringement being a fair use (depending on the other three factors).

The Court has potentially opened a floodgate of possible changes, in this case, depending on how the lower courts interpret the opinion. Congress has tinkered with copyright law since the advent of the internet (which changed everything) but has not engaged in a comprehensive revision since 1976. So, what do we suspect that the implications of this case may be?

Now that the line is brighter if not clear between copyright and patent protection, a prudent software company will rely increasingly on patent protection to protect the heart of the work and less on default copyright assumptions that predate this case.

Using patent law may be futile since the Supreme Court and the Federal Circuit have left patent lawyers scratching their heads after Alice Corp. v. CLS Bank International regarding the future of software patents. Congress’s last major revision, the Patent Act in 2011, was unhelpful. Maybe the answer should be found in a future revision to the Copyright Act sharpening the line between patent and copyright protection of software.

Courts also have historically ignored the infringer’s positive externalities to the copyright holder — when the consumption or production of a good causes a benefit to a third party. However, now, potentially, if there is infringement, but the infringement adds value, does it create a stronger argument for a finding of fair use? After all, courts often give greater weight to the effect on the market than to the other factors. So, are we on a slow course to a “use it or lose it” copyright regime? That’s a position the United States has rejected in the international copyright context. 

Now must a copyright owner exploit, try to exploit, or plan to exploit the copyrighted work in all possible markets, or risk having another do so? Only the future will tell, but the answer can come sooner than once in a generation of transformational Supreme Court decisions. 

If Congress exercises its prerogatives and once again considers transformational and comprehensive revisions to the Copyright Act that carefully balance the interests of creators and users of copyrighted materials, then maybe there will be a settled law of intellectual property in the software context.

Llewellyn Joseph Gibbons is a distinguished university professor of law at The University of Toledo College of Law.