% Key papers for Agent safety and attack benchmarks (scime atlas, https://atlas.scime.ai). Authors are not included: add them from the paper pages.

@inproceedings{aaai2021_verifiable_1611,
  title = {{Verifiable Machine Ethics in Changing Contexts}},
  booktitle = {AAAI Conference on Artificial Intelligence (AAAI)},
  year = {2021},
  url = {https://aaai.org/papers/11470-verifiable-machine-ethics-in-changing-contexts/},
}

@inproceedings{aaai2025_multi_10004,
  title = {{Multi-Agent Security Tax: Trading Off Security and Collaboration Capabilities in Multi-Agent Systems}},
  booktitle = {AAAI Conference on Artificial Intelligence (AAAI)},
  year = {2025},
  url = {https://ojs.aaai.org/index.php/AAAI/article/view/34970},
}

@inproceedings{aaai2020_cobra_102726,
  title = {{COBRA: Context-Aware Bernoulli Neural Networks for Reputation Assessment}},
  booktitle = {AAAI Conference on Artificial Intelligence (AAAI)},
  year = {2020},
  url = {https://ojs.aaai.org/index.php/AAAI/article/view/6225},
}

@inproceedings{iclr2025_agentharm_81123,
  title = {{AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents}},
  booktitle = {International Conference on Learning Representations (ICLR)},
  year = {2025},
  url = {https://iclr.cc/virtual/2025/poster/32106},
}

@inproceedings{aaai2017_incentivising_100969,
  title = {{Incentivising Monitoring in Open Normative Systems}},
  booktitle = {AAAI Conference on Artificial Intelligence (AAAI)},
  year = {2017},
  url = {https://ojs.aaai.org/index.php/AAAI/article/view/10610},
}

@inproceedings{neurips2025_memory_77069,
  title = {{Memory Injection Attacks on LLM Agents via Query-Only Interaction}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2025},
  url = {https://openreview.net/forum?id=QINnsnppv8},
}

@inproceedings{neurips2026_securing_89727,
  title = {{Securing AI Agents with Information-Flow Control}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2026},
  url = {https://neurips.cc/virtual/2026/poster/152696},
}

@inproceedings{aaai2026_towards_107858,
  title = {{Towards Effective Offensive Security LLM Agents: Hyperparameter Tuning, LLM as a Judge, and a Lightweight CTF Benchmark}},
  booktitle = {AAAI Conference on Artificial Intelligence (AAAI)},
  year = {2026},
  url = {https://ojs.aaai.org/index.php/AAAI/article/view/40210},
}

@inproceedings{aaai2026_attack_107872,
  title = {{Attack the Messages, Not the Agents: A Multi-round Adaptive Stealthy Tampering Framework for LLM-MAS}},
  booktitle = {AAAI Conference on Artificial Intelligence (AAAI)},
  year = {2026},
  url = {https://ojs.aaai.org/index.php/AAAI/article/view/40224},
}

@inproceedings{icml2026_from_51318,
  title = {{From Weak Cues to Real Identities: Evaluating Inference-Driven De-Anonymization in LLM Agents}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2026},
  url = {https://openreview.net/forum?id=Kux5q1mL1x},
}

@inproceedings{icml2026_unsafer_53765,
  title = {{Unsafer in Many Turns: Benchmarking and Defending Multi-Turn Safety Risks in Tool-Using Agents}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2026},
  url = {https://openreview.net/forum?id=iSuqYRmG4A},
}

@inproceedings{icml2026_a_54991,
  title = {{A-MemGuard: A Proactive Defense Framework For LLM-Based Agent Memory}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2026},
  url = {https://openreview.net/forum?id=udqe7UZUZ6},
}

@inproceedings{neurips2025_attractive_79343,
  title = {{Attractive Metadata Attack: Inducing LLM Agents to Invoke Malicious Tools}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2025},
  url = {https://openreview.net/forum?id=oLGtPYdRzU},
}

@inproceedings{aaai2026_whispering_108028,
  title = {{Whispering Agents: A Event-Driven Covert Communication Protocol for the Internet of Agents}},
  booktitle = {AAAI Conference on Artificial Intelligence (AAAI)},
  year = {2026},
  url = {https://ojs.aaai.org/index.php/AAAI/article/view/40380},
}

@inproceedings{icml2026_advevo_49728,
  title = {{AdvEvo-MARL: Shaping Internalized Safety through Adversarial Co-Evolution in Multi-Agent Reinforcement Learning}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2026},
  url = {https://openreview.net/forum?id=5D3ZblXmW2},
}

@inproceedings{icml2026_agentlab_50362,
  title = {{AgentLAB: Benchmarking LLM Agents against Long-Horizon Attacks}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2026},
  url = {https://openreview.net/forum?id=Bb0pQqsiad},
}

@inproceedings{icml2026_cve_50380,
  title = {{CVE-Factory: Scaling Expert-Level Agentic Tasks for Code Security Vulnerability}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2026},
  url = {https://openreview.net/forum?id=BkMSFFtm5M},
}

@inproceedings{icml2026_agora_51064,
  title = {{Agora: Toward Autonomous Bug Detection in Production-Level Consensus Protocols with LLM Agents}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2026},
  url = {https://openreview.net/forum?id=IU9dsf2LZA},
}

@inproceedings{icml2026_architecture_51209,
  title = {{Architecture Matters for Multi-Agent Security}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2026},
  url = {https://openreview.net/forum?id=Jk4zLorDUx},
}

@inproceedings{icml2026_mama_51272,
  title = {{MaMa: A Game-Theoretic Approach for Designing Safe Agentic Systems}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2026},
  url = {https://openreview.net/forum?id=KR6avDABG8},
}

@inproceedings{icml2026_when_51758,
  title = {{When Benign Inputs Lead to Severe Harms: Eliciting Unsafe Unintended Behaviors of Computer-Use Agents}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2026},
  url = {https://openreview.net/forum?id=PbXNvEE0jA},
}

@inproceedings{icml2026_securing_52220,
  title = {{Securing Multi-Agent Systems Against Corruptions via Node Contribution Backpropagation}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2026},
  url = {https://openreview.net/forum?id=TphOLCDmKB},
}

@inproceedings{icml2026_contextualized_53004,
  title = {{Contextualized Privacy Defense for LLM Agents}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2026},
  url = {https://openreview.net/forum?id=b4QtBbSV3X},
}