Prof. Doug Lichtman, UCLA School of Law
August 22, 2011
Google’s Android operating system is under siege. Oracle has a substantial patent and copyright case underway in which it alleges that Android is illegally derivative of the Sun programming language, Java. Apple has literally dozens of patent cases underway, alleging that Samsung, HTC, and Motorola in various ways infringe Apple’s intellectual property whenever those firms produce Android-capable handsets. Microsoft has its own Android litigations underway, including most interestingly an International Trade Commission case against Motorola, Google’s would-be bride. And Google in response has been making headlines of its own, primarily by spending and offering to spend unimaginable sums of money to acquire patents from entities like Nortel, IBM, and (according to the rumors) InterDigital and Kodak.
I have a lot to say about all those overlapping and disheartening legal maneuvers, but here I want to focus on yet another threat to Android that is supposedly looming: a challenge that would undermine Google’s ability to use and in other ways incorporate Linux source code into the Android ecosystem. Here’s the background.
Android is built in part on source code that was originally written for Linux, the open-source operating system. Google was well within its rights to use that source code; the code was released under what is known as the GPLv2 license, which is a license that explicitly authorizes reuse inside other software ventures. The GPLv2 license, however, comes with some strings. Among them, anyone who reuses the code within a new, larger project has to disclose, publicly, certain parts of the resulting code. Google and its partners have done some such disclosure but apparently have not disclosed enough to satisfy that license term.
A large number of technology thinkers and intellectual property lawyers have noticed this shortcoming. The law firm of Brown Rudnick sounded the alarm earlier this month in a series comprised of Part 1 and Part 2 on the firm’s site. Commentator Florian Mueller similarly called attention to the issue this month, flagging the concern on his uniformly thoughtful intellectual property blog, Foss Patents. Even the Free Software Foundation, the nonprofit organization that wrote the original GPLv2 license, has in the last few days gone public with this same basic story.
What happens if Google really did make this mistake? The three sites I mention above paint a dire picture: Having failed to comply in the past, Google and its partners have lost their right to use the Linux code; coming into compliance now would in no way reinstate that prior license; and thus Android is today vulnerable to copyright suits from each of the many individual authors who contributed relevant code to the open-source project.
Gathering permission from all of those authors, of course, would be all but impossible. The code at issue was presumably written by dozens or maybe even hundreds of individual authors. It is not today clear who those authors are, nor is it clear how Google would go about contacting them. Worse, each author would have a strong incentive to withhold permission if contacted, holding hostage the Android ecosystem and ultimately demanding a ransom disproportionate to that author’s own original contribution.
But I am not yet ready to panic. After all, if the GPLv2 license really does terminate in the event of an error like the one Google is accused of, then any disagreement between Google and the original authors of the relevant Linux code will, by default, be governed by federal copyright law. Among other things, that means that Google can defend its alleged unauthorized copying by arguing fair use.
Will that help? I think so. Fair use is at its heart an equitable defense, and – assuming that Google’s mistake here was an innocent one and further assuming that Google is moving quickly to rectify it now that the error has been pointed out – the equities here would strongly favor Google. For one thing, any complaining copyright holder would have a hard time showing real “factor four” economic harm; those authors gave their work away for free and did so with the very intent of seeing it reused in other programming efforts. Moreover, the fact that it would at this point be difficult to seek out permission from the dozens of relevant authors also argues in favor of fair use. Indeed, one of the standard justifications for the fair use doctrine is that it can excuse infringement in cases where licensing is impractical.
Overall, then, I doubt that this newest problem for Android will in the end cause any major disruptions for Google and its partners. Yes, the incident does point out yet another latent defect in the GPLv2 approach. And, as I note here, the incident might also provide an interesting opportunity for a court to think carefully about how fair use ought to apply to open source licenses. But I suspect that this fight, at least, will ultimately resolve in Google’s favor.
Open source licensing is supposed to be a driver of innovation. The law ought not, and likely will not, allow its increasingly burdensome complexity to undermine that important goal.