Securing Multi-Agent Systems Against Corruptions via Node Contribution Backpropagation
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866910259441827840 |
|---|---|
| author | Wu, Chengcan Zhang, Zhixin Xu, Mingqian Wei, Zeming Sun, Meng |
| author_facet | Wu, Chengcan Zhang, Zhixin Xu, Mingqian Wei, Zeming Sun, Meng |
| contents | Multi-Agent Systems (MAS) have become a prevalent paradigm for Large Language Model (LLM) applications. However, the complex multi-agent design in MAS introduces unique trustworthiness concerns: adversarial agents can inject misleading information that propagates contagiously through the system, corrupting benign agents and leading to false outputs. Existing graph-based defenses model agents as nodes and communications as edges, yet are limited to static-graph defenses. In this paper, we propose a dynamic defense paradigm that models MAS communication as a signed directed acyclic graph and computes each agent's contribution to the final decision via backward propagation, enabling accurate identification and isolation of malicious agents to secure multi-agent task collaboration. Experimental results in complex and dynamic MAS environments demonstrate that our method notably outperforms existing MAS defense mechanisms, providing an effective guardrail for trustworthy MAS deployment. Our code is available at https://github.com/ChengcanWu/BPD. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_19420 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Securing Multi-Agent Systems Against Corruptions via Node Contribution Backpropagation Wu, Chengcan Zhang, Zhixin Xu, Mingqian Wei, Zeming Sun, Meng Cryptography and Security Artificial Intelligence Machine Learning Multiagent Systems Optimization and Control Multi-Agent Systems (MAS) have become a prevalent paradigm for Large Language Model (LLM) applications. However, the complex multi-agent design in MAS introduces unique trustworthiness concerns: adversarial agents can inject misleading information that propagates contagiously through the system, corrupting benign agents and leading to false outputs. Existing graph-based defenses model agents as nodes and communications as edges, yet are limited to static-graph defenses. In this paper, we propose a dynamic defense paradigm that models MAS communication as a signed directed acyclic graph and computes each agent's contribution to the final decision via backward propagation, enabling accurate identification and isolation of malicious agents to secure multi-agent task collaboration. Experimental results in complex and dynamic MAS environments demonstrate that our method notably outperforms existing MAS defense mechanisms, providing an effective guardrail for trustworthy MAS deployment. Our code is available at https://github.com/ChengcanWu/BPD. |
| title | Securing Multi-Agent Systems Against Corruptions via Node Contribution Backpropagation |
| topic | Cryptography and Security Artificial Intelligence Machine Learning Multiagent Systems Optimization and Control |
| url | https://arxiv.org/abs/2510.19420 |