Mitigating Generative AI Hallucinations
(A2026-133661)

AUTHORS

Shubhranshu Singh, Johns Hopkins University

ABSTRACT

We theoretically investigate whether AI developers or AI operators should be liable for the harm the AI systems may cause when they hallucinate, i.e., when they generate confident and plausible but false outputs. We develop a novel theoretical framework to study and compare implications of liability for developers of AI systems versus liability for human decision makers. We find that the optimal liability framework may vary over time, with the evolution of the AI technology, and that making the AI operators liable can be desirable only if it induces monitoring of the AI systems. We also highlight non-trivial relationships between welfare and reputational concerns, human supervision ability, and the accuracy of the technology. Our results have implications for regulatory design and business strategies.