A privileged identity is a high-trust account or service identity that can do more than a normal user, such as manage systems, access sensitive data, or call protected APIs. Examples include administrators, automation accounts, and AI agents that are allowed to act on behalf of a person or platform. Because these identities carry broad permissions, they also have a larger blast radius if stolen, misconfigured, or abused.
In cyber security, privileged identities are a primary target for attackers because compromise can lead to rapid lateral movement, data theft, or destructive changes. Defenders reduce risk with least privilege, short-lived and scoped credentials, multifactor authentication, separation of duties, and strong audit trails. Monitoring unusual token use, unexpected tool calls, or access outside normal patterns can reveal abuse early. Treating software agents and service accounts like privileged identities helps organizations control automation without giving it unchecked power.