Private Law Theory

ABSTRACT
When an AI system trained on copyrighted data in the United States, deployed by an Indian user, and hosted on a Chinese platform generates a deepfake that defames a French citizen, which jurisdiction’s law governs the liability? This paper argues that the answer is currently ‘none’. The rapid proliferation of AI-generated content (AIGC) has exposed a fundamental fracture in private international law. Existing frameworks are not merely divergent; they are epistemically incompatible. The European Union’s precautionary strict liability, the United States’ market-driven immunity (Section 230), China’s state-supervised control, and India’s statutory vacuum create a ‘Compliance Cliff’ where simultaneous adherence to all regimes is mathematically impossible. Through a comparative analysis of these four jurisdictions, this article demonstrates how this fragmentation facilitates regulatory arbitrage, leaving victims of algorithmic harm, from defamation to copyright infringement, without a functional remedy. To resolve this deadlock, the paper proposes a ‘Tiered Value-Chain Accountability’ framework. Rather than seeking a utopian global treaty, it advances operational mechanisms, including a ‘Rome II Plus 2’ standard for jurisdiction, a Voluntary Licensing Clearinghouse (WAIL), and ISO-standard watermarking, to bridge incompatible regimes. By shifting liability from ‘territoriality’ to ‘demonstrable control’, this framework offers a pathway to crossborder enforcement that respects jurisdictional diversity while closing the immunity gaps that currently shield algorithmic harm.

Pathak, Shashwat, Private Law Challenges of AI-Generated Content in Cross-Border Digital Markets (February 1, 2026).