Zero Standing Privilege
Zero standing privilege is an access-control principle under which no identity holds persistent, always-on permissions. Instead of granting a user or service a standing role that is available at any time, access is issued just-in-time for a specific task and expires when the task is done. It is a core idea in modern zero-trust design, aimed at shrinking the window in which a compromised credential is useful. Applied to AI agents, zero standing privilege means an agent does not carry broad, permanent authority to call models or tools. Each action is authorized in the moment, against policy, rather than assumed from a standing grant.
How zero standing privilege changes the default
The common pattern gives an account a role and leaves it in place: the permissions sit there whether or not they are being used, which is exactly what an attacker inherits when they steal the credential. Zero standing privilege inverts the default to deny, then grants narrow access for a specific operation and revokes it after. The credential a thief captures at rest carries little usable authority, because authority attaches to authorized actions rather than to the identity's resting state.
How it maps to AI agents
An AI agent with a broad, standing token can call any model or tool that token permits, at any time, which is what makes a hijacked agent dangerous. Enforcing zero standing privilege for agents means evaluating each request when it is made rather than trusting a persistent grant. DeepInspect authorizes AI traffic per request against per-role and per-route policy tied to the verified identity, so an agent's ability to reach a given model or tool depends on the policy verdict at that moment. The per-decision audit record then shows each grant as an event, not as an assumption.
Related reading
- AI Agent Privilege Scoping: Six Patterns That Contain an Agent's Blast Radius
An agent is a program that acts on behalf of a human, and the acting has authorization consequences the traditional privilege model does not cover. The agent's identity, the human's session, the tool's permission, and the enterprise policy all compose into the authorization decision on each call. Privilege scoping is the design pattern set that keeps the composed authorization tight. This piece walks through six patterns that appear in production agent deployments and the audit records each pattern produces.
- Zero Trust AI: Per-Request Evaluation at the Model Boundary
Zero trust applied to AI means evaluating every model request against verified identity, current policy, and prompt-level classification. The architectural pattern is an enforcement proxy at the HTTP AI request boundary. The post-authentication gap is the most common failure mode in current deployments.