Agentic identity is the practice of giving AI agents, software that can act autonomously on a person's or organization's behalf, their own verifiable identity, rather than treating them as an extension of a human session or a shared API key. It extends identity and access management beyond people to include the software that increasingly acts for them.
Frequently asked questions
Why can't an AI agent just reuse the customer's session?
A shared session can't distinguish what the human did from what the agent did, which breaks accountability and makes it impossible to scope or revoke the agent's access independently.
Is agentic identity a separate product from CIAM?
No, it's CIAM extended to cover a new identity type; see agentic CIAM.
What standards does agentic identity build on?
The same OAuth 2.0 and OpenID Connect foundation as everything else in this glossary, with extensions like scoped access and delegated consent layered on top.
.avif)