paperJune 2026Unreviewed

Securing the AI Agent: A Unified Framework for Multi-Layer Agent Red Teaming

Yong Yang, Xing Zheng, Huiyu Wu, Huangsheng Cheng, Xiaorong Shi, Jing Guo, Bo Yang, Yi Zhou, Xiangfan Wu, Zonghao Ying

Abstract

The fast growth of open-source AI infrastructure, from model serving engines and agent platforms to the Model Context Protocol (MCP) ecosystem and the language models themselves, has outpaced the security tooling available to defend it. We present AI-Infra-Guard, an open-source framework that organizes AI red teaming around a single observation: the attack surface of an AI agent is stratified across layers (infrastructure, protocol/tool, agent behavior, and model), and no single detection paradi

Framework mappings

NIST AI Risk Management Framework

MEASUREMeasure

Suggested from the entry's categories.

Cite

@misc{yang2026securing,
  title = {{Securing the AI Agent: A Unified Framework for Multi-Layer Agent Red Teaming}},
  author = {Yong Yang and Xing Zheng and Huiyu Wu and Huangsheng Cheng and Xiaorong Shi and Jing Guo and Bo Yang and Yi Zhou and Xiangfan Wu and Zonghao Ying},
  year = {2026},
  month = jun,
  eprint = {2606.31227},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2606.31227}
}