Preprint · 2026 · Open access

CoSec: Benchmarking Agent Security in Communities

Hao Mark Chen, Wenhui Dong, Ye Chen, Jiezhi Yao, Chenbo Xia, 曲玉文, Renxiang Wang, 袁复栋, Camil Hamami, Chenglong Pan, Xinquan Yue, Ziyu Wang and 3 more

arXiv (Cornell University) · Published 28 September 2026 · DOI 10.48550/arxiv.2609.34790

Open PDFView at the publisherCite this paper

In brief

“CoSec: Benchmarking Agent Security in Communities” is a 2026 preprint by Hao Mark Chen and 14 co-authors, published in arXiv (Cornell University). OpenAlex counts 1 citations as of 19 November 2026. It is open access under a CC BY licence: the full text is free to read at arXiv (Cornell University).

Citations
1
Most cited in
2026 (1)
References
24
Authors
15

Abstract

LLM agents operate in persistent collaborative environments involving multiple users, communities, memories, files, and tools. Community boundaries may remain fixed or evolve with changes in membership, roles, composition, and relationships. Agents must complete legitimate tasks and prevent unauthorized disclosure of protected information. Existing evaluations do not fully examine these risks in agent systems. We introduce \textbf{CoSec}, an executable benchmark for evaluating privacy and authorization enforcement in LLM agent systems operating within and across communities. CoSec contains 208 canonical scenarios spanning fixed and evolving boundaries, protected information belonging to the agent owner or other participants, and attacks through dialogue, environmental content, persistent memory, and composed workflows. CoSec executes complete agent systems with persistent sessions, memory, files and tools. It verifies information flows against the active authorization state using execution traces and artifacts. Across harness and model configurations, agents frequently complete benign tasks but violate privacy and authorization boundaries. Privacy behavior varies across harnesses, attack surfaces, and community states, revealing how memory, files, tools, and workflows can carry protected information beyond its authorized scope. These findings show that task utility does not imply privacy or authorization compliance and that authorization in community settings remains an unresolved security challenge for persistent LLM agents.

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Chen, H. M., Dong, W., Chen, Y., Yao, J., Xia, C., 曲玉文, Wang, R., 袁复栋, Hamami, C., Pan, C., Yue, X., Wang, Z., Ye, F., Si, C., & Shan, C. (2026). CoSec: Benchmarking Agent Security in Communities. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2609.34790

MLA 9

Chen, Hao Mark, et al. “CoSec: Benchmarking Agent Security in Communities.” arXiv (Cornell University), 2026. https://doi.org/10.48550/arxiv.2609.34790.

Chicago

Chen, Hao Mark, Wenhui Dong, Ye Chen, Jiezhi Yao, Chenbo Xia, 曲玉文, Renxiang Wang, et al. 2026. “CoSec: Benchmarking Agent Security in Communities.” arXiv (Cornell University). https://doi.org/10.48550/arxiv.2609.34790.

Harvard

Chen, H. M. et al. (2026) ‘CoSec: Benchmarking Agent Security in Communities’, arXiv (Cornell University). doi:10.48550/arxiv.2609.34790.

IEEE

H. M. Chen et al., “CoSec: Benchmarking Agent Security in Communities,” arXiv (Cornell University), 2026, doi: 10.48550/arxiv.2609.34790.

BibTeX

@misc{chen2026cosec,
  title     = {{CoSec: Benchmarking Agent Security in Communities}},
  author    = {Chen, Hao Mark and Dong, Wenhui and Chen, Ye and Yao, Jiezhi and Xia, Chenbo and 曲玉文 and Wang, Renxiang and 袁复栋 and Hamami, Camil and Pan, Chenglong and Yue, Xinquan and Wang, Ziyu and Ye, Fengyu and Si, Chenyang and Shan, Caifeng},
  howpublished = {arXiv (Cornell University)},
  year      = {2026},
  doi       = {10.48550/arxiv.2609.34790},
  url       = {https://doi.org/10.48550/arxiv.2609.34790}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Topic and keywords

Physical Sciences › Computer Science › Computer Networks and Communications › Access Control and Trust

  • LLM agents
  • agent security
  • privacy enforcement
  • benchmarking
  • persistent memory
  • information leakage
  • community membership
  • agent cooperation

Work with this paper

Most-cited papers in Computer Science

All highly cited Computer Science papers

Questions about this paper

How do I cite this paper in APA style?

Chen, H. M., Dong, W., Chen, Y., Yao, J., Xia, C., 曲玉文, Wang, R., 袁复栋, Hamami, C., Pan, C., Yue, X., Wang, Z., Ye, F., Si, C., & Shan, C. (2026). CoSec: Benchmarking Agent Security in Communities. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2609.34790

How many times has this paper been cited?

1 times, according to OpenAlex on 19 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.

Is this paper free to read?

Yes. It is open access at arXiv (Cornell University): sign in and open the PDF from this page.

Who wrote this paper?

Hao Mark Chen, Wenhui Dong, Ye Chen, Jiezhi Yao, Chenbo Xia, 曲玉文, Renxiang Wang, 袁复栋, Camil Hamami, Chenglong Pan, Xinquan Yue, Ziyu Wang and 3 more.

What is the DOI of this paper?

10.48550/arxiv.2609.34790, permanently at https://doi.org/10.48550/arxiv.2609.34790. Sign in and open it from this page.

Bibliographic data from OpenAlex (CC0), last checked 19 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.