arXiv cs.CL· Mohammad Kachuee, Phu Mon Htut, Sheng Li, Nikolaos Pappas, Miguel Ballesteros Martinez, Yassine Benajiba, Sujith Ravi, Dan Roth·· 12 小时前AI 评分42
MAS-Guard：从单个智能体到多智能体 LLM 系统的统一防护框架
MAS-Guard: Unified Guardrails from Individual Agents to Multi-Agent LLM Systems
AI 导读
研究者提出统一防护框架 MAS-Guard，覆盖单个与多智能体 LLM 应用，在 6 个基准、3 个不同厂商的智能体 LLM 上测试，相对无防护智能体将整体攻击成功率降低 71.5-86.3%，同时在良性任务上满足效用参考标准。
正文 · 原文
Abstract:Agentic AI extends large language models (LLMs) from generating responses to planning and executing complex workflows, substantially expanding their utility while exposing applications to unsafe actions, information leakage, and adversarial manipulation. Multi-agent systems further expand these capabilities through specialization and collaboration, but distributed authority, heterogeneous roles, and opaque or potentially compromised participants make safeguarding even more challenging. Existing defenses address important individual threats but provide fragmented coverage across local execution, system-level interactions, and enforcement mechanisms. We introduce MAS-Guard, a unified framework that closes this gap for safeguarding single- and multi-agent AI applications containing both instrumentable and opaque agents. Within instrumentable agents, the proposed local Agent Guardrail (A-Guard) layer applies task-specific experts to assess agentic events. At the system level, developer-owned Agent Charters and a Control Plane combine deterministic structural checks of routes, agent capabilities, and message formats with semantic assessments of roles and message content. These complementary layers enable enforcement through a shared set of interventions: allow, reflection, confirmation, and block. Evaluated across six benchmarks covering diverse single- and multi-agent scenarios with three agent LLMs from different providers, MAS-Guard reduces the overall attack success rate by 71.5-86.3% relative to unguarded agents, while meeting the utility reference criterion on benign tasks.
| Subjects: | Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.28397 [cs.CR] |
| (or arXiv:2610.28397v1 [cs.CR] for this version) | |
| https://doi.org/10.48550/arXiv.2610.28397 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Mohammad Kachuee Mr. [view email]
[v1]
Tue, 20 Oct 2026 15:44:37 UTC (3,360 KB)
来源：arXiv cs.CL · arxiv.org