Large World Models (LWMs)— the next generation of AI systems capable of generating persistent, interactive, physics-coherent worlds—are poised to unsettle long-standing assumptions across multiple areas of law.
This program examines the emerging challenges as simulated environments become embedded in commercial platforms, professional decision-making, and public systems. We will explore tort liability arising from reliance on simulated worlds and world-scale forecasts; new contract difficulties when LWMs function as dynamic, evolving counterparties; novel copyright questions in synthetic environments populated with AI-generated assets and derivative works; and evidentiary issues related to the authentication, admissibility, and epistemic status of LWM-generated reconstructions. The session also addresses one of the most profound questions these systems raise—whether highly autonomous agents embedded in large world models require new categories of legal personhood or responsibility. Participants will gain a grounded understanding of how LWMs challenge traditional legal doctrine and where the next generation of rules must evolve.