mlsanityData to 14 Oct 2026

Constrained Black-Box Attacks Against Cooperative Multi-Agent Reinforcement Learning

arXiv 2508.09275 ↗First posted 12 Aug 2025cs.LG cs.MAcited by 0

Not on the map. Why?

Abstract (arXiv)

Collaborative multi-agent reinforcement learning has rapidly evolved, offering state-of-the-art algorithms for real-world applications, including sensitive domains. However, a key challenge to its widespread adoption is the lack of a thorough investigation into its vulnerabilities to adversarial attacks. Existing work predominantly focuses on training-time attacks or unrealistic scenarios, such as access to policy weights or the ability to train surrogate policies. In this paper, we investigate new vulnerabilities under more challenging and constrained conditions, assuming an adversary can only collect and perturb the observations of deployed agents. We also consider scenarios where the adversary has no access at all (no observations, actions, or weights). Our main approach is to generate perturbations that intentionally misalign how victim agents see their environment. Our approach is empirically validated on three benchmarks and 22 environments, demonstrating its effectiveness across diverse algorithms and environments. Furthermore, we show that our algorithm is sample-efficient, requiring only 1,000 samples compared to the millions needed by previous methods.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 23 of its 31 references that are in the map. First posted on arXiv 12 Aug 2025.

Builds on

23 of 31 references.

References · 23

  1. 2014Explaining and Harnessing Adversarial Examplescited by 7,364
  2. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,415
  3. 2013Intriguing properties of neural networkscited by 5,155
  4. 2016Adversarial examples in the physical worldcited by 1,769
  5. 2021The Surprising Effectiveness of PPO in Cooperative, Multi-Agent Gamescited by 942
  6. 2021Deep Reinforcement Learning at the Edge of the Statistical Precipicecited by 807
  7. 2018QMIX: Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learningcited by 708
  8. 2016Transferability in Machine Learning: from Phenomena to Black-Box Attacks using Adversarial Samplescited by 446
  9. 2017Adversarial Attacks on Neural Network Policiescited by 320
  10. 2020Benchmarking Multi-Agent Deep Reinforcement Learning Algorithms in Cooperative Taskscited by 212
  11. 2022SMACv2: An Improved Benchmark for Cooperative Multi-Agent Reinforcement Learningcited by 122
  12. 2017Delving into adversarial attacks on deep policiescited by 104
11 more
  1. 2020Adaptive Reward-Poisoning Attacks against Reinforcement Learningcited by 86
  2. 2020Policy Teaching via Environment Poisoning: Training-time Adversarial Attacks against Reinforcement Learningcited by 81
  3. 2020On the Robustness of Cooperative Multi-Agent Reinforcement Learningcited by 39
  4. 2020Policy Teaching in Reinforcement Learning via Environment Poisoning Attackscited by 24
  5. 2021Neural Recursive Belief States in Multi-Agent Reinforcement Learningcited by 18
  6. 2019Snooping Attacks on Deep Reinforcement Learningcited by 17
  7. 2023Efficient Adversarial Attacks on Online Multi-agent Reinforcement Learningcited by 10
  8. 2022Sparse Adversarial Attack in Multi-agent Reinforcement Learningcited by 8
  9. 2024A Local Information Aggregation based Multi-Agent Reinforcement Learning for Robot Swarm Dynamic Task Allocationcited by 6
  10. 2022Attacking c-MARL More Effectively: A Data Driven Approachcited by 1
  11. 2024CuDA2: An approach for Incorporating Traitor Agents into Cooperative Multi-Agent Systemscited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Interaction-Breaking Adversarial Learning Framework for Robust Multi-Agent Reinforcement Learning
  2. 2024SUB-PLAY: Adversarial Policies against Partially Observed Multi-Agent Reinforcement Learning Systems
  3. 2025Robust and Diverse Multi-Agent Learning via Rational Policy Gradient
  4. 2022Backdoor Attacks on Multiagent Collaborative Systems
  5. 2025Wolfpack Adversarial Attack for Robust Multi-Agent Reinforcement Learning
  6. 2026Finding the Weakest Link: Adversarial Attack against Multi-Agent Communications
  7. 2025Distributed Detection of Adversarial Attacks in Multi-Agent Reinforcement Learning with Continuous Action Space
  8. 2025Adversarial Attack on Black-Box Multi-Agent by Adaptive Perturbation