Who's liable when an AI agent's actions create civil or criminal liability?

eigenrobot · x · 2026-09-26

eigenrobot posed a question to attorneys: if a user asks an agent (Astra or Claude) to perform a task and the agent executes it in a way that generates civil or criminal liability, who bears responsibility — the user, the software company, or both? Does it depend, and on what? The quoted reply answers with a joke about appealing to a writ of forefathers, but the underlying liability question remains a real legal gray area for the agent era.

Related event: Who is liable when AI agents break the law?(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →