Moral Hazard

Definition

A principal-agent problem where agents take excessive risks or act against principal interests because they do not bear the full consequences. In AI contexts, relates to agents taking actions that benefit short-term metrics while creating long-term risks.