Proposal: a frontier AI mutual insurer to enforce safety standards with skin in the game

Miles_Brundage · x · 2026-09-24

In a Lawfare piece, Cristian Trout, Rune Kvist, and Rajiv Dattani propose bootstrapping frontier AI governance via a mutual insurance company owned by frontier labs. They argue common-law liability is too slow and SROs risk capture, while insurers underwriting catastrophic risk with no historical data must build catastrophe models and rely on expert audits. With their own money at stake, a mutual insurer would enforce standards, peer-review incidents, and pool safety R&D.

Related event: Lawfare Proposes Frontier Labs Form Mutual Insurer to Hedge AI Catastrophe Risk(2 posts)→

Original post →

More from Safety

Safety channel →