Copyrightability of APIs: a small skirmish in a much larger war

The Electronic Frontier Foundation is launching a new effort ("No Copyright on APIs: Help Us Make the Case") to fight the copyrightability of APIs at the appellate level, presumably as an amicus brief participant in the Oracle v. Google appeal.

My question is, what's the point of fighting copyrightability of APIs if those same APIs can be readily protected by specification licenses or terms of use? The Amazon Web Services API is freely available because Amazon's terms say so. If Amazon decided that certain use of the API is no longer authorized (say, in order to lock out competitors), Amazon need only change the terms.  At a minimum, Amazon has a contract claim against violators, perhaps not as powerful as a copyright infringement claim, but a claim nonetheless and sufficiently chilling.

All of this is happening despite the substantial precedent (relied on by Judge Allsup) that has - since the 90s - casted serious doubt on the copyrightability of APIs. Google gambled wisely on that issue when it moved ahead with Dalvik on Android, and ultimately won decisively despite Oracle's heroic efforts at trial to secure copyright protection for Java platform APIs. But even without Judge Allsup's decision, I think most software lawyers (all the ones I know, anyway) are in general agreement: although the law in this area is not completely settled, the weight of authority is on the side of no protection for APIs. And yet, Twitter changes the terms of access to its APIs all the time, and entire swaths of competitors get wiped out as a result.

An issue that is just as important as the copyrightability of APIs is how copyrightable content and contract intersect. With these take-it-or-leave-it shrinkwrap and clickwrap agreements, content becomes walled off to the public, in ways that betray the balance the US Copyright Act strives to achieve: a balance between proprietary rights and public access. With the proliferation of mandatory non-negotiable terms enveloping more and more content, creating a new property right via contract, how can we protect fair use rights? How can we protect access to what should be public domain or uncopyrightable material? And in the context of software, how can we preserve interoperability, competition and innovation? In short, how can we prevent the death of copyright?

I think it would be worth exploring changes to the Copyright Act to address these licenses. One modest proposal I have is if a license or set of terms abrogates reverse engineering or fair use rights, or attempts to assert control over content that is not protected by copyright, the license should contain a warning of this fact explicitly, at the top of the license in conspicuous bold text – akin to a consumer protection measure. And, if the licensor chooses to abrogate such rights, then the licensor's remedies should be limited to contract remedies if there's a breach of the license. A licensor of content imposing unilateral terms stripping fair use rights should not be permitted to recover copyright remedies if that license is breached. Failure to comply with the conspicuous labelling requirement would also lead to preclusion of ocpyright remedies.

This alone would not sufficiently address the issue, but it would raise public awareness and eventually perhaps public opinion.

notices    log in

© 2006, 2017 Sean Hogle PC. All rights reserved. These pages constitute ATTORNEY ADVERTISEMENT.