Both Donald Trump and Bernie Sanders want to nationalize AI. So do many others, from bipartisan congressional committees to national security hawks to academic critical theorists. Should governments nationalize frontier AI? Part of the problem is that no one knows what nationalization means: the leading proposals are quite different from one another. This talk makes two contributions. First, it defines nationalization. Drawing on the economic theory of ownership, nationalization bundles two separable entitlements: residual claim rights, or who captures the surplus, and residual control rights, or who directs the actions that no contract or statute anticipates. Existing plans target totally different rights. Second, the talk argues for a form of minimalist nationalization, involving control rights rather than claim rights, and covering just two dangers from frontier AI: catastrophic risk and hard corporate power. Inspired by European golden shares, the government could halt a frontier AI company’s training or deployment of a catastrophically risky system. For other risks, including national security failures, monopoly power and inequality, ordinary regulation and taxation are better options. The talk is based on joint work with Yonathan Arbel and Peter Salib.

About the Speaker:
Simon Goldstein is Associate Professor of Philosophy at the University of Hong Kong and a Senior Editor at AI Frontiers. He works on the ethics and law of AI: whether AIs have interests of their own, what legal standing they should have, and how humanity can avoid catastrophe as AI grows more capable. He is the author of three books: AI Rights (with Peter Salib, Cambridge University Press), AI Welfare: Agency, Consciousness, Sentience (with Cameron Domenico Kirk-Giannini, Oxford University Press), and Iterated Knowledge (Oxford University Press). He received his PhD from Rutgers.

Register here to attend the seminar on or before 6 October 2026, 12:00noon (Hong Kong Time).