paperMay 2026Unreviewed

Agent Security is a Systems Problem

Mihai Christodorescu, Earlence Fernandes, Ashish Hooda, Somesh Jha, J. Rehberger, Kamalika Chaudhuri, Xiaohan Fu, Khawaja Shams, Guy Amir, Jihye Choi, Sarthak Choudhary, Nils Palumbo, Andrey Labunets, Nishit V. Pandya

arXiv.org

View resource

Abstract

We take the position that agent security must be approached as a systems problem: the AI model powering the agent must be treated as an untrusted component, and security invariants must be enforced at the system level. Through this lens, efforts to increase model robustness (the dominant viewpoint in the community) are insufficient on their own. Instead, we must complement existing efforts with techniques from the systems security domain. Based on our experience as cybersecurity researchers in o

Cite

@misc{christodorescu2026agent,
  title = {{Agent Security is a Systems Problem}},
  author = {Mihai Christodorescu and Earlence Fernandes and Ashish Hooda and Somesh Jha and J. Rehberger and Kamalika Chaudhuri and Xiaohan Fu and Khawaja Shams and Guy Amir and Jihye Choi and Sarthak Choudhary and Nils Palumbo and Andrey Labunets and Nishit V. Pandya},
  year = {2026},
  month = may,
  eprint = {2605.18991},
  archivePrefix = {arXiv},
  doi = {10.48550/arXiv.2605.18991},
  url = {https://www.semanticscholar.org/paper/44d178dd44370773c5d0536025c8c848febcee9b}
}