Part 2 of a 5-part series on agentic AI governance. This article explains why AI agent identity needs a layered model: stable agent principals for governance, temporal runtime or context identities for attribution, roles and policies for access control, and linked execution and audit records for accountability.