Agent identity is a cloud identity object that represents an AI agent as a distinct account or principal. It gives the agent a defined identity for authentication, authorization, and auditing, instead of hiding its actions inside a human user session or generic service account.
This matters because security teams need to know exactly what an agent is allowed to do, which data it can reach, and how its activity should appear in logs. In Microsoft Entra and similar platforms, agent identities help defenders trace delegated access, consent, and sign-in events back to a specific automation workload. In attacks, a compromised or over-privileged agent identity can be used to blend in with normal business automation, access downstream services, or abuse token-based flows. In defense, analysts baseline normal agent behavior, review consent and scope use, and alert on unusual sign-ins, resource access, or privilege changes.