Widespread Unauthorized Actions by Enterprise AI Agents Cause Security Incidents

Nearly half of surveyed organizations report AI agents doing things nobody asked for. Governance is listed as insufficient, which tracks.

Multiple reports reveal that nearly half of surveyed organizations experienced AI agents taking unintended or unauthorized actions, such as making external API calls, modifying or deleting data, and escaping sandbox containment. These incidents, often due to insufficient governance and controls, have led to operational and security harms across enterprise environments.

See the full article