Identity and Access Management refers to the systems that authenticate users or services, grant them permissions, and enforce access policies across enterprise or cloud environments. IAM covers user directories, service accounts, role-based permissions, single sign-on, and federated identity across multiple systems.
Every major cloud provider has a native IAM system — AWS IAM, Google Cloud IAM, Azure Active Directory (now Entra ID). Enterprise environments typically combine cloud-native IAM with on-premise directory services (Active Directory) via federation protocols (SAML, OIDC).
IAM has become a load-bearing concern for AI agent deployment. Autonomous agents operate with service accounts that must be carefully scoped — too little permission and the agent cannot do its job; too much permission and a misbehaving agent can cause material damage. The principle of least privilege, long an IAM best practice, has become a production-critical discipline for enterprise agent deployment.