MAS-Guard: Unified Guardrails from Individual Agents to Multi-Agent LLM Systems

原文：MAS-Guard: Unified Guardrails from Individual Agents to Multi-Agent LLM Systems

作者：Mohammad Kachuee, Phu Mon Htut, Sheng Li, Nikolaos Pappas, Miguel Ballesteros Martinez, Yassine Benajiba, Sujith Ravi, Dan Roth

来源：arXiv cs.AI（人工智能）

正文

Computer Science > Cryptography and Security

arXiv:2610.28397v1 (cs)

[Submitted on 20 Oct 2026]

Title:MAS-Guard: Unified Guardrails from Individual Agents to Multi-Agent LLM Systems

Authors:Mohammad Kachuee, Phu Mon Htut, Sheng Li, Nikolaos Pappas, Miguel Ballesteros Martinez, Yassine Benajiba, Sujith Ravi, Dan Roth

View a PDF of the paper titled MAS-Guard: Unified Guardrails from Individual Agents to Multi-Agent LLM Systems, by Mohammad Kachuee and 7 other authors

View PDF

HTML (experimental)

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

Focus to learn more

arXiv-issued DOI via DataCite

主题

安全 · 人工智能 · 自然语言处理

由「前沿雷达」于 2026-10-20 采集。正文取自原文页面，已保留出处链接。标题与正文版权归原作者所有。