Deny Without Disabling: Authorization-Paired Evaluation and Control for Multi-Agent Systems

2026-09-30

AI Summary

This study addresses the risk in multi-agent collaboration where individually compliant information, when combined, may lead to unauthorized use. To mitigate this, we propose an authorization pairing evaluation mechanism and the FlowReview framework, which leverage object parsing, permission ordering, and deterministic execution control to prevent illicit information composition without disrupting legitimate cooperation. This work innovatively establishes “prohibiting violations” and “completing lawful tasks” as joint success criteria, thereby balancing security with system efficacy. Experimental results demonstrate that the proposed approach reduces the submission rejection rate from 86% to zero while preserving full authorized provisioning capability, effectively resolving the challenge of multi-agent coordination under stringent security constraints.

Abstract

Multi-agent systems derive their capabilities from sharing evidence, delegating tasks, and combining information across agents. The same process creates a safety problem: contributions that are admissible in isolation can jointly enable a prohibited use. Blocking every sensitive action avoids disclosure but defeats the purpose of collaboration. We introduce authorization-paired evaluation, which makes blocking prohibited uses and completing required authorized uses a joint success criterion, and FlowReview, a framework connecting object resolution, permission ranking, and deterministic enforcement. In controlled composition experiments, reviewing combined artifacts reduces the denied-commit rate from 86.0% to zero with no loss of authorized supply. Our findings show that preserving information and lineage alone does not ensure correct permission attribution. Object identity and permission must remain connected to execution through components whose outputs can be verified. Together, these findings establish a system-level requirement for multi-agent safety: govern composed information flows while preserving the authorized capabilities that make collaboration useful.

Problem

• Multi-Agent Systems
• Safety
• Authorization
• Information Flow Control
• Composed Artifacts

Innovation

• Authorization-Paired Evaluation
• FlowReview
• Multi-Agent Systems
• Deterministic Enforcement
• Composed Information Flows