Intent-Governed Tool Authorization for AI Agents
2026-06-22 • Artificial Intelligence
Artificial Intelligence
AI summaryⓘ
The authors explain that current ways of letting AI agents use external tools only check if the agent has permission, but not if the action matches what the user actually wants. They propose a new system called Intent-Governed Access Control (IGAC) that adds a layer making sure tool use fits the user's intent and limits what the AI can do. IGAC creates special intent certificates and checks so that permissions can only be narrowed, never expanded, based on what the user requests. This system is integrated with an existing governance tool called OpenPort to improve security and accountability.
AI agentsauthorizationIntent-Governed Access Control (IGAC)intent certificatespolicy narrowingmanifest filteringpayload consistencyOpenPortauthorization-dependent discoveryattribute-based access control (ABAC)
Authors
Genliang Zhu, Chu Wang
Abstract
AI agents increasingly act through external tools: they read private data, construct structured payloads, submit write requests, export records, and coordinate workflows across application boundaries. Existing authorization mechanisms usually ask whether an integration credential, app, or token can call a tool. That question is necessary but incomplete. A tool call can be authorized by static credentials and still be unjustified by the user's current request. For example, a credential that can read and export records should not expose export authority when the user only asked for a bounded summary, and a model-generated delete call should not execute merely because the integration has a delete scope. This paper proposes Intent-Governed Access Control (IGAC), a server-side authorization layer that treats the user's expressed intent as a monotone, auditable policy attribute for AI-agent tool use. IGAC introduces intent certificates, session-scoped policy narrowing, intent-aware manifest filtering, and intent-tool-payload consistency checks. The central invariant is that user intent may only reduce the authority granted by static integration policy; it never expands scopes, data policy, tenant boundaries, or review requirements. We map IGAC onto OpenPort, an existing governance substrate that already implements authorization-dependent discovery, scope and ABAC-style policy checks, draft-first writes, preflight impact binding, state-witness checks, idempotency, stable reason codes, and audit.