Earlier this month the Court of Appeals for the Federal Circuit issued the latest ruling in the long-running legal battle between two Silicon Valley giants Oracle, the hardware and software maker, and Google. Oracle claimed that Google’s Android operating system infringes on patents and copyrights it possessed in its Java operating system. A lower court had previously found that there was no patent infringement, a finding that was unaffected by this latest ruling. The lower court had also found that Google had not infringed on any Oracle copyrights in files associated with the Java API (more on what this is in a bit) on the grounds that, according to copyright law, Oracle had no copyright in the API that could be infringed. The appeals court, however, has disagreed, thereby unleashing an uproar over the damaging effects that the ruling, if allowed to stand, poses for future innovation.
Why such consternation over what seems to be an arcane bit of law connected to an esoteric bit of technology? In part because the technology is not so esoteric. APIs, or application programming interfaces, are part of nearly every technology, particularly computer technology. They are what enable different technologies to work together.
For instance, if you want to print this webpage, you will click “print” in your browser. But your browser needs to know how to talk to your computer, and your computer needs to know how to talk to your printer in order for the page to be printed. APIs articulate how these instructions can be passed from one technology to another. Thus anyone who wants his or her technology to work with any other technology is going to need to use APIs.
Concern has arisen because copyright law grants creators of original content a great deal of control over how others use it. However, because the control it offers creators comes at a cost to the public who would seek to use what they’ve created, the law also contains limitations regarding the sort of content that can be protected by copyright. The question before the court was whether the kind of content APIs represent is the type that is copyrightable. Up until this ruling the answer appeared to be no.
Nor are APIs the kind of content that smart innovation policy would want to be eligible for copyright protection. Even though, as some have argued, it takes some effort to produce a good API, as the Supreme Court has noted in several previous decisions, the primary objective of copyright is not to reward the labor of authors but, as the Constitution instructs in Article I, Section 8, “to promote the progress of science and useful arts.”
Full disclosure: I’ve done legal work for Google, although not on this case. But this ruling isn’t about Google or Android or even Java; it’s about the damage it stands to cause anyone trying to innovate. By design, copyrights are monopolies, and we tolerate them only insofar as doing so is necessary to promote progress. But rulings like this one that extend the reach of copyright do the opposite and risk transforming copyright into a barrier to progress instead. The implications of this ruling transcend the interests of Oracle and Google. It stands to affect any technology (present and future) that needs to work with another — which, effectively, is all of them. The upshot of this ruling is that according to it, anyone who owns a copyright in an API can say no to anyone whose innovation requires it — a situation that’s simply not good for progress or the public that depends on it.
By either misunderstanding what sort of content those API files represented, or misreading the statute and applicable precedent that should have applied to them, the ruling effectively places limits on how others can develop their new technologies. It means the copyright owner of an API in a technology can stand as a gatekeeper, either extracting a tax on others who want to make sure their own technology works with it, or even block them entirely. Which is exactly what Oracle tried to do with this lawsuit. By letting it, the appeals court has now invited anyone else who has developed an API to use copyright to do the same, even if it comes at the expense of innovation or a competitive marketplace.
The court’s mistake is thus not harmless error. Nor is it even necessary.
Proponents of expansive views of copyright often argue that without it, no one would have any incentive to create. But this is a contention worth questioning. For one, as the saying goes, if you build a better mousetrap, the world will beat a path to your door. You don’t need the exclusive monopoly granted by copyright to eliminate competition if you have the better product. For another, this case shows how irrelevant being able to copyright APIs has been in encouraging their development. Plenty of people have been developing them for years without the expectation (given existing statutory language and judicial precedent) that the APIs would be copyrightable.
Furthermore, to the extent that technology developers are incentivized to innovate only by the prospect of profiting from their technology’s being widely used, being able to lock down APIs (the means for other technologies to work with them) only serves to impede that goal. Unless, of course, the actual goal is to profit through the elimination of competition; you don’t need to build a better mousetrap if yours is the only one the public can buy.
In fact, rather than incentivizing innovation, allowing copyright in APIs only serves to disincentivize it, for both the copyright holder (who now doesn't need to produce the best product in the market) and other innovators and developers, because it exposes them to potentially significant legal risk and expense if they need to comport with an API to make their technologies work properly.
Technically, this ruling is not the final word in this case. Even if it is not overturned by subsequent appeal, the court left open the question of whether Google’s use of the APIs was permissible under the doctrine of fair use. But having to hope that fair use might provide a defense for copyright infringement creates a chilling amount of uncertainty for the innovator. As noted copyright academic Lawrence Lessig famously observed, “Fair use is the right to hire a lawyer.” In other words, the costs to innovate dramatically expand when you need to defend what you’ve developed on the basis of fair use. It is much better for all who depend on APIs to innovate for them not to be copyrightable at all, as there will never be a need to defend one’s use against claims of infringement for something that couldn’t be infringed in the first place.