When is software speech protected by the First Amendment?

November 23rd, 2012

Earlier this year, Google commissioned Eugene Volokh to author a white paper laying out the position that information generated by algorithms, namely search results, are entitled to First Amendment protections. The Times picked up the story. Then, Google also commissioned a paper from Robert Bork (!) and Gregory Sidak arguing that Google’s policies are consistent with antitrust rules. These moves are a preemptive response to pending investigations by the FTC that Google’s search results violate antitrust rules. In other words, advancing defenses based on the First Amendment defense, as well as antitrust law.

The Stanford Law Review Online has an interesting piece by Yale 3L Andrew Tutt, titled “Software Speech” looking at whether information should receive First Amendment protection, in light of Brown v. EMA and Sorrell v. IMS Health. Here is the crux of his short piece:

 To enact a sound information policy, the Court should neither embrace a seemingly absurd result (as Sorrell would counsel) nor look to narrow analogies (as Brown would counsel) and instead look to the broader and more difficult question of the degree to which a class or category of new media implicates the First Amendment’s core purposes. Rather than counseling greater protection from governmental interference, this may in fact suggest that the government have a freer hand in content-neutral software regulation. Software is sometimes primarily concerned with conveying ideas of the kind and in a manner that one would recognize as familiar and essential to a free society. At other times, software functions much more like a means by which data is gathered, manipulated, and relayed to and by a user and therefore difficult to think of as akin to “speech.” . . .

The Court can similarly create a category of “software speech” granting all software a modicum of heightened scrutiny given the important First Amendment concerns its regulation might raise. But by stopping short of calling software “speech,” entirely and unequivocally, the Court would acknowledge the many ways in which software is still an evolving cultural phenomenon unlike others that have come before it. In discarding tests for whether software is speech on the basis of its literal resemblance either to storytelling (Brown) or information dissemination (Sorrell), the Court would strike a careful balance between the legitimate need to regulate software, on the one hand, and the need to protect ideas and viewpoints from manipulation and suppression, on the other.

I am working on an article, tentatively tiled Robot, Esq., that looks at a specific type of software speech–algorithms that can generate legal advice, and the First Amendment implications of such technology (see some posts I wrote on this topic earlier this year here).

I think this line of reasoning will become quite relevant in the near future. Google search results are just the beginning.