Jason Jia-Xi Wu*1
[This essay is available in PDF at this link]
Abstract
The United States and China are engaged in a regulatory arms race over artificial intelligence (AI). Yet, existing debates often overlook a critical factor shaping this AI race: federalism, or the division of regulatory authority between central and local governments. In the United States, states lead in AI regulation, with the federal government taking a limited, backseat role. In China, although authority remains more centralized, local governments have played a pivotal role in implementing and experimenting with AI policies. While institutional differences remain, both countries exhibit signs of partial convergence towards a fragmentary approach to AI governance.
What explains this convergence? The Article argues that the answer lies in industry self-governance. In both countries, the tech industry is increasingly acting as a co-regulator of AI systems alongside traditional central and local authorities. As gatekeepers, suppliers, and beneficiaries of disruptive AI technologies, the tech industry imposes market discipline on regulators at both levels, often by leveraging local protectionism and jurisdictional variation to advance its interests. However, as national security takes center stage in this AI race, the tech industry is assuming both commercial and geopolitical roles, emerging as a third regulatory force that reshapes center-local relations.
This new paradigm reflects what this Article terms “techno-federalism.” Blending “technocracy” with “federalism,” techno-federalism captures how emerging AI norms both disrupt and transform center-local relations by empowering the private sector to embed itself in public policymaking. It challenges the dominant view that the U.S.-China AI race is merely a “battle of values” between liberal democracy and techno-autocracy. By highlighting the tripartite interplay between central, local, and market power, techno-federalism offers a more nuanced perspective, addressing the limits of conventional geostrategic approaches to the U.S.-China AI race.
*1 Judicial Law Clerk, U.S. District Court; J.D., Harvard Law School; M.A., Harvard Univ.
To Valentina—a constant compass, a quiet grace. Your kindness lives in these pages.
I am grateful to Margaret Siu, whose excellent research assistance enriched my work immeasurably. I also thank Dr. Guanchi Zhang, Dr. Ying Wu, Zeming Liu, Wenda Xiang, Chentuo Zhu, and Hounong Li for their insightful comments. Finally, I am indebted to Kristle Hinola, Sonia Mahajan, William Yee, and the editorial board of the Harvard National Security Journal for their meticulous editorial support and tireless efforts. The views expressed here are my own and do not reflect those of the U.S. District Court. All errors and omissions are solely my responsibility.