---
title: "Assurance-Conditioned Continuous Authorization for Autonomous AI Agents"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2611.05632v1"
permalink: "https://feed.astralor.com/items/itm_7e5fcc9e58082b28337b886b"
published: "2026-11-03T21:19:46.000Z"
score: 59
channel: "ai"
tags: ["AI Agent","Assurance-Conditioned Continuous Authorization","Autonomous","Agents Autonomous","ACCA","Protected"]
---

# Assurance-Conditioned Continuous Authorization for Autonomous AI Agents

## 速览

Autonomous software agents can act through tools, delegate work, and change their models or execution environments while earlier authorizations remain in flight. We present Assurance-Conditioned Continuous Authorization…

## 关键点

- Autonomous software agents can act through tools, delegate work, and change their models or execution environments while earlier authorizations remain in flight.
- We present Assurance-Conditioned Continuous Authorization (ACCA), an approach designed to be harness-independent, for maintaining operational authority according to currently justified assurance.
- It distinguishes evidence validity from contextual applicability, derives assurance claims from eligible evidence, and resolves a multidimensional effective-authority envelope within a principal-declared ceiling and app…

## 全文

Autonomous software agents can act through tools, delegate work, and change their models or execution environments while earlier authorizations remain in flight. We present Assurance-Conditioned Continuous Authorization (ACCA), an approach designed to be harness-independent, for maintaining operational authority according to currently justified assurance. It distinguishes evidence validity from contextual applicability, derives assurance claims from eligible evidence, and resolves a multidimensional effective-authority envelope within a principal-declared ceiling and applicable policy. A typed dependency graph supports selective re-evaluation of authorization-relevant changes. An authority epoch changes when effective authority changes; a distinct optional authorization-basis version supports revalidation of action-specific authorization conditions. Protected effect-time enforcement can reject stale actions before a consequential resource change, subject to the available commit mechanism. A compositional Domain Governance Profile expresses domain-specific semantics without granting authority. Controlled scenario tests implemented as a regression suite using a trading profile exercise default-deny rules, evidence-conditioned contraction and recovery, policy divergence, selective non-response, and stale-action rejection at a local gateway before the simulated-broker boundary. The study establishes implementation-level evidence in one domain, not a cross-domain guarantee of harmful-behavior prevention.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_7e5fcc9e58082b28337b886b