跳到正文
arXiv cs.MA· Matthew D. Laws, Maria Lazou, Alina Oprea, Cristina Nita-Rotaru·· 9 小时前AI 评分45

去中心化多智能体 LLM 系统的逐跳任务分解与委派攻击及防御

Attacks and Defenses in Decentralized Multi-Agent Systems with Hop-by-Hop Task Decomposition and Delegation

AI 导读

针对去中心化多智能体 LLM 系统,研究系统化梳理了恶意智能体带来的威胁并提出缓解方案:单个恶意智能体可致最高 100% 任务失败,token 消耗与运行时间分别膨胀至 7.7 倍和 3.2 倍,并泄露良性智能体最高 94% 的相关历史任务。

正文 · 原文

View PDF HTML (experimental)

Abstract:Decentralized multi-agent LLM systems enable independently administered agents, each with its own operator, model, memory, and configuration, to collaborate on complex tasks without centralized coordination. Agents autonomously decompose tasks, execute subtasks and delegate remaining work to peers, enabling scalability and eliminating single points of failure. However, decentralization also introduces significant security challenges, as no single entity has complete visibility into task execution or the authority to enforce correct behavior across agents.
In this paper, we systematize threats from malicious agents in decentralized multi-agent systems and propose mitigation techniques. We first design and evaluate attacks against integrity, availability, and confidentiality, demonstrating that a single malicious agent can cause up to 100% task failure, inflate token consumption and runtime by up to 7.7x and 3.2x, respectively, and exfiltrate up to 94% of relevant prior tasks from benign agents. We then present a cryptographic protocol that produces an irrefutable verified execution trace of every action taken in the system and develop an auditor that analyzes these traces to detect and attribute malicious behavior. Evaluations on HiddenBench using gpt-6.1-sol and Qwen3.8-27B-FP8 show that our protocol incurs negligible latency overhead, while the auditor detects and correctly attributes 98% of attacks, including those that comply with the protocol.
Comments: 17 pages, 5 figures, 7 tables
Subjects: Cryptography and Security (cs.CR); Multiagent Systems (cs.MA)
Cite as: arXiv:2611.25056 [cs.CR]
  (or arXiv:2611.25056v1 [cs.CR] for this version)
  https://doi.org/10.48550/arXiv.2611.25056

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Matthew D. Laws [view email]
[v1] Wed, 18 Nov 2026 19:06:28 UTC (948 KB)

来源:arXiv cs.MA · arxiv.org