Abstract

The rise of artificial intelligence (AI) and cross-border data flows has exposed significant doctrinal, institutional, and geopolitical gaps in the multilateral trading system, especially acutely in the U.S.-China trade conflict. This paper examines how AI and data are currently regulated under the WTO framework, specifically GATS, TRIPS, and the SCM Agreement, and analyses how divergent models of digital capitalism in the United States and China create structural tensions over state-owned enterprises, industrial subsidies, and national security exceptions. Drawing on doctrinal analysis of WTO jurisprudence and comparative treatment of regional instruments such as the CPTPP, USMCA, and RCEP, the paper argues that plurilateral soft-law undertakings within the WTO offer a viable legal mechanism to bridge regulatory divergence without demanding treaty amendment or regulatory convergence. The paper further contends that the viability of such undertakings depends on three complementary institutional features: non-adjudicatory oversight mechanisms, substantive engagement with conflicting conceptions of digital sovereignty, and mandatory sunset review clauses. Absent such institutional reform, soft-law undertakings risk becoming aspirational statements that legitimize continued friendshoring and unilateral action, accelerating fragmentation of the multilateral trading system.

This work is licensed under a Creative Commons Attribution 4.0 License.