The principle of least privilege, often abbreviated PoLP, limits identities to the minimum permissions and resources required. Narrow access reduces the harm that mistakes, compromised credentials or unintended behavior can cause.
Agent permissions must remain useful enough to complete approved work, so least privilege is not simply the smallest possible token. Task-specific identity, temporary elevation and action budgets can provide capability without granting unrestricted access.
Acronyms and aliases
PoLP acronymleast privilege variant
Related terms
Frequently asked questions
Why is least privilege important for AI agents?
Agents can act quickly and repeatedly, so unnecessary permissions can turn one error into widespread or irreversible damage.
Does least privilege make agents unable to work?
It should not. Permissions should match the approved task, with controlled escalation when additional authority is genuinely needed.
