Answers
What is agent identity, and who is liable when an agent acts?
The answer
Agent identity is the record of which agent took an action, under whose authority, on whose behalf. Without it an organization cannot attribute a commitment, and an unattributable commitment is a liability question with no defendant — which is why identity is the first control, not a later one.
The technical form is familiar: a durable identifier, a declared set of capabilities, scoped credentials, and a signature on every action. The consequential part is the attribution chain behind it. When an agent commits the firm to something, three facts must be recoverable a year later — which agent, which policy permitted it, and which human owned that policy at the time. Systems that log only the action leave every review to be argued from memory.
Across organizational boundaries the question sharpens. An agent transacting with another company’s agent needs to prove what it is and what it is allowed to ask for, before any work happens. That is a discovery and trust problem rather than a language problem, and it is the reason identity is the precondition for agent-to-agent communication, for permissions, and for anything resembling agent commerce.