Definition
A proposed legal structure that would let an AI agent run a corporation while a human principal is shielded from liability for the agent's actions. The idea has been taken up in a state proposal and drawn a congressional investigation over who is accountable when the agent acts.
Why it matters
If an agent can legally own and operate a business, the question of who pays when it signs a bad contract or breaks a law becomes critical for insurers, investors, and regulators.