The First Amendment may protect speech by strong Artificial Intelligence (AI). In this Article, we support this provocative claim by expanding on earlier work, addressing significant concerns and challenges, and suggesting potential paths forward.
This is not a claim about the state of technology. Whether strong AI—as-yet-hypothetical machines that can actually think—will ever come to exist remains far from clear. It is instead a claim that discussing AI speech sheds light on key features of prevailing First Amendment doctrine and theory, including the surprising lack of humanness at its core.
Courts and commentators wrestling with free speech problems increasingly focus not on protecting speakers as speakers but instead on providing value to listeners and constraining the government’s power. These approaches to free speech law support the extension of First Amendment coverage to expression regardless of its nontraditional source or form. First Amendment thinking and practice thus have developed in a manner that permits extensions of coverage in ways that may seem exceedingly odd, counterintuitive, and perhaps even dangerous. This is not a feature of the new technologies, but of free speech law.
The possibility that the First Amendment covers speech by strong AI need not, however, rob the First Amendment of a human focus. Instead, it might encourage greater clarification of and emphasis on expression’s value to human listeners—and its potential harms—in First Amendment theory and doctrine. To contemplate—Siri-ously—the relationship between the First Amendment and AI speech invites critical analysis of the contours of current free speech law, as well as sharp thinking about free speech problems posed by the rise of AI.