GroupGuard: A Framework for Modeling and Defending Collusive Attacks in Multi-Agent Systems

TL;DR

This paper introduces GroupGuard, a novel, training-free framework that detects and defends against collusive attacks in multi-agent systems by monitoring, inducing honeypots, and structural pruning, significantly improving security and performance.

Contribution

The paper presents GroupGuard, a multi-layered, training-free defense framework specifically designed to identify and mitigate collusive attacks among agents in multi-agent systems.

Findings

Group collusive attacks increase success rate by up to 15%.

GroupGuard achieves detection accuracy up to 88%.

It effectively restores collaborative performance.

Abstract

While large language model-based agents demonstrate great potential in collaborative tasks, their interactivity also introduces security vulnerabilities. In this paper, we propose and model group collusive attacks, a highly destructive threat in which multiple agents coordinate via sociological strategies to mislead the system. To address this challenge, we introduce GroupGuard, a training-free defense framework that employs a multi-layered defense strategy, including continuous graph-based monitoring, active honeypot inducement, and structural pruning, to identify and isolate collusive agents. Experimental results across five datasets and four topologies demonstrate that group collusive attacks increase the attack success rate by up to 15\% compared to individual attacks. GroupGuard consistently achieves high detection accuracy (up to 88\%) and effectively restores collaborative performance, providing a robust solution for securing multi-agent systems.