mlsanityData to 14 Oct 2026

Finding the Weakest Link: Adversarial Attack against Multi-Agent Communications

arXiv 2605.13170 ↗First posted 13 May 2026cs.LG cs.MAcited by 0

Not on the map. Why?

Abstract (arXiv)

Multi-agent systems rely on communication for information sharing and action coordination, which exposes a vulnerability to attacks. We investigate single-victim communication perturbation attacks against Multi-Agent Reinforcement Learning-trained systems and propose methods that use gradient information from the Jacobian to identify which messages, agent, and timesteps are most susceptible to attack and have the greatest impact on the system. We enhance these methods with two proposed adversarial loss functions that trade-off attack success for attack impact which also create more effective perturbations. We empirically demonstrate the effectiveness of our methods against two different multi-agent communication methods in navigation, PredatorPrey, and TrafficJunction environments. Our results show that our novel message selection method achieves a similar or greater impact than random message selection across almost all tested scenarios. Our victim selection, message selection, tempo, and loss functions improve attack effectiveness in half of the thirty scenarios we tested.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 19 of its 32 references that are in the map. First posted on arXiv 13 May 2026.

Builds on

19 of 32 references.

References · 19

  1. 2014Explaining and Harnessing Adversarial Examplescited by 7,364
  2. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,415
  3. 2013Playing Atari with Deep Reinforcement Learningcited by 3,274
  4. 2015The Limitations of Deep Learning in Adversarial Settingscited by 790
  5. 2016Learning Multiagent Communication with Backpropagationcited by 366
  6. 2017Adversarial Attacks on Neural Network Policiescited by 320
  7. 2017Tactics of Adversarial Attack on Deep Reinforcement Learning Agentscited by 166
  8. 2017Robust Deep Reinforcement Learning with Adversarial Attackscited by 154
  9. 2018Learning when to Communicate at Scale in Multiagent Cooperative and Competitive Taskscited by 121
  10. 2017Delving into adversarial attacks on deep policiescited by 104
  11. 2022A Survey of Multi-Agent Deep Reinforcement Learning with Communicationcited by 104
  12. 2020Stealthy and Efficient Adversarial Attacks against Deep Reinforcement Learningcited by 63
7 more
  1. 2021Who Is the Strongest Enemy? Towards Optimal and Efficient Evasion Attacks in Deep RLcited by 55
  2. 2021Adversarial Attacks On Multi-Agent Communicationcited by 49
  3. 2020Succinct and Robust Multi-Agent Communication With Temporal Message Controlcited by 29
  4. 2021Mis-spoke or mis-lead: Achieving Robustness in Multi-Agent Communicative Reinforcement Learningcited by 18
  5. 2021Understanding Adversarial Attacks on Observations in Deep Reinforcement Learningcited by 13
  6. 2017Revisiting the Master-Slave Architecture in Multi-Agent Deep Reinforcement Learningcited by 12
  7. 2018Multi-agent Deep Reinforcement Learning with Extremely Noisy Observationscited by 10

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Constrained Black-Box Attacks Against Cooperative Multi-Agent Reinforcement Learning
  2. 2026Interaction-Breaking Adversarial Learning Framework for Robust Multi-Agent Reinforcement Learning
  3. 2026Safety of Latent Communication in Multi-Agent Systems
  4. 2025textitAgents Under Siege: Breaking Pragmatic Multi-Agent LLM Systems with Optimized Prompt Attacks
  5. 2025Multi-Agent Reinforcement Learning with Communication-Constrained Priors
  6. 2025Adversarial Attack on Black-Box Multi-Agent by Adaptive Perturbation
  7. 2025Red-Teaming LLM Multi-Agent Systems via Communication Attacks
  8. 2025Exposing Weak Links in Multi-Agent Systems under Adversarial Prompting