paperSeptember 2026Unreviewed

SoK: When Safe Agents Fail Together: The Security of Multi Agent LLM Systems

Rui Yang, Junjie Xu, Zhengyu Liu, Neil Fendley, Yang Hong, Ziyang Li, Yinzhi Cao

Abstract

Safe agents can fail together. Multi-agent LLM systems (MAS) move information, state, decisions, and authority across principal boundaries, creating failures that local checks may miss. Without an execution-level view, a multi-agent setting can easily be mistaken for evidence of a genuinely multi-agent security effect. We thus systematize MAS security through an execution-centered analysis of 197 works, covering six interaction interfaces, four adversary positions, seven system-level risks, and

Cite

@misc{yang2026sok,
  title = {{SoK: When Safe Agents Fail Together: The Security of Multi Agent LLM Systems}},
  author = {Rui Yang and Junjie Xu and Zhengyu Liu and Neil Fendley and Yang Hong and Ziyang Li and Yinzhi Cao},
  year = {2026},
  month = sep,
  eprint = {2609.00595},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2609.00595}
}