Agent identity
The property that an autonomous agent is distinguishable as a specific actor — with a durable identifier, a recorded authority grant, and an attributable action history — rather than operating under a shared service credential.
The position
Agents are still overwhelmingly deployed as generic service accounts, and that single fact defeats most of the audit apparatus built around them. If three agents share a credential, the question 'which one acted' has no answer at the identity layer, and every downstream reconstruction is inference. This is the least glamorous prerequisite in agentic governance and the one that has to come first: authority cannot be granted to something that cannot be named, and accountability cannot attach to something that cannot be distinguished.
This section is a claim, not a definition. It is argued rather than asserted, and it is the part of this page you are invited to disagree with. Corrections and counter-arguments to [email protected] are published.