CFP: Disciplining AI

As the world increasingly runs on AI, humanity has yet to discover how best to curb AI harms. Current legal policy and scholarship emphasize two strategies. The first imposes liability on people after their AI causes harm. The second regulates AI more directly—its development and its deployment—but before harm occurs. “Disciplining AI” will explore a third approach: directly intervening with AI after it causes harm. Co-hosted by the Harvard Journal of Law and Technology and Iowa’s Innovation, Business, and Law Center, symposiasts will convene for an in-person event on April 2, 2027, at Harvard Law School. They will consider the possibility and advisability of applying sanctions, or sanction-like responses, to AI.
We welcome all methodologies and all fields of study. Perspectives may constructively or critically engage any step of the disciplinary process, from articulating liability standards, to detecting violations, to crafting interventions. HJOLT is particularly interested in submissions that engage one of the following four subthemes:  

1. Corporate Law and Personhood: Must AI be a legal person for the law to discipline it? What sort of legal person would AI be? Could existing legal structures facilitate the introduction of AI legal personhood, e.g. by creating corporations that are co-extensive with an AI system?
2. Theories of Punishment: What model of discipline is best suited to AI? Does or could AI have the sort of “motivation” needed for deterrence to work? Is it possible to “rehabilitate” AI? Could targeted incapacitation prevent future offenses?
3. Civil Liberties: What Constitutional protections does AI discipline implicate? Free speech? Due process? Takings? Would these safeguards protect AI itself or those who develop and deploy it?
4. Implementation: Which jurisdiction has the power or the authority to discipline AI? Do we need an international framework? Within the United States, which agencies are relevant? What role could there be for non-state actors, including market forces, private coordination, or the technical research community?

  Authors of accepted proposals will prepare essays (maximum 8,000 words) for presentation at the symposium and subsequent publication in Volume 40 of HJOLT. Symposiasts will each receive an honorarium of $1750 to defray the costs of travel and accommodations. Invited symposiasts include Ryan Abbott (Surrey), Joanna Bryson (Hertie School), Pauline Kim (WashU), Frank Pasquale (Cornell), Alexander Sarch (Surrey), and Jordi Weinstock (Harvard).   To apply to “Disciplining AI,” please submit an abstract (300-500 words) and CV to [email protected], with CC to [email protected], by October 26, 2026. Abstracts should clearly indicate which (if any) of the four subthemes above they engage.