ABSTRACT
We study optimal liability for AI-powered products with semi-autonomous capabilities (eg, self-driving vehicles). Like human users, AI can inadvertently cause product failures that harm third parties. Additionally, AI may introduce an extreme risk of large-scale social harm that renders full liability impractical. Raising AI liability for ordinary loss above the actual harm level can decrease excessive autonomy and increase social welfare, even when it negatively impacts R&D efforts. A well-designed liability rule can implement efficient levels of autonomy and balanced R&D that reduces AI’s general risk, but full efficiency cannot be achieved under targeted R&D that reduces AI’s extreme risk.
Chen, Yongmin and Hua, Xinyu, Product Safety in the Age of AI: Autonomy, R&D, and AI Liability (July 16, 2024), HKUST Business School Research Paper No 2024-158.
Leave a Reply