Robin Hanson's "Foom Liability": a robust policy case for AI liability incentives

NathanpmYoung · x · 2026-09-14

The thread discusses Robin Hanson's essay "Foom Liability," which argues for robust AI policy: instead of betting on specific doom-risk estimates, impose liability incentives on AI builders so that society benefits if foom risk is high while costs stay low if it isn't. Hanson reasons that near-miss variants of the self-improving AI scenario are far more numerous than the extreme case. The poster calls for an organization lobbying specifically for AI liability.

Original post →

More from AGI Musings

AGI Musings channel →