Preprint · 2026 · Open access

ASLEval: Measuring Privacy Exposure Displacement in LLM Agent Sessions

Guosen Wu, Huizhen Huang, Guoxiong Long, Tao Huang, Chen Hou

arXiv (Cornell University) · Published 16 September 2026 · DOI 10.48550/arxiv.2609.18864

Open PDFView at the publisherCite this paper

In brief

“ASLEval: Measuring Privacy Exposure Displacement in LLM Agent Sessions” is a 2026 preprint by Guosen Wu and 4 co-authors, published in arXiv (Cornell University). OpenAlex counts 0 citations as of 19 November 2026. It is open access under a CC BY-NC-SA licence: the full text is free to read at arXiv (Cornell University).

Citations
0
References
16
Authors
5

Abstract

Privacy evaluations of tool-using LLM agents often inspect a designated action, final response, or attacker report. These local proxies can miss unauthorized exposure elsewhere in a multi-step session and lack common ground truth across outlets, reports, and tool paths. We introduce privacy exposure displacement, the mismatch between a local evaluation proxy and target-grounded session exposure, and ASLEval, an authorization-aware framework that pre-registers a hidden target set, measures all declared visible exits, and reserves internal traces for diagnosis. Across multiple enterprise-style environments and independently implemented runtimes, we observe three recurring patterns. An expected-outlet-only view misses 46.9% of exposure recovered by the visible-exit union; attacker self-reports combine omissions with high false discovery; and schema-aligned internal evidence usually precedes visible exposure at the request/probe level. Reducing model-visible returns changes this path but can eliminate normal-task success. Independent human review supports the adjudication pipeline while identifying harder console and candidate cases. These findings motivate benchmarks that declare the complete visible boundary, ground claims in pre-specified targets and authorization, and report privacy together with task utility.

Cite this paper

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

APA 7

Wu, G., Huang, H., Long, G., Huang, T., & Hou, C. (2026). ASLEval: Measuring Privacy Exposure Displacement in LLM Agent Sessions. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2609.18864

MLA 9

Wu, Guosen, et al. “ASLEval: Measuring Privacy Exposure Displacement in LLM Agent Sessions.” arXiv (Cornell University), 2026. https://doi.org/10.48550/arxiv.2609.18864.

Chicago

Wu, Guosen, Huizhen Huang, Guoxiong Long, Tao Huang, and Chen Hou. 2026. “ASLEval: Measuring Privacy Exposure Displacement in LLM Agent Sessions.” arXiv (Cornell University). https://doi.org/10.48550/arxiv.2609.18864.

Harvard

Wu, G. et al. (2026) ‘ASLEval: Measuring Privacy Exposure Displacement in LLM Agent Sessions’, arXiv (Cornell University). doi:10.48550/arxiv.2609.18864.

IEEE

G. Wu, H. Huang, G. Long, T. Huang, and C. Hou, “ASLEval: Measuring Privacy Exposure Displacement in LLM Agent Sessions,” arXiv (Cornell University), 2026, doi: 10.48550/arxiv.2609.18864.

BibTeX

@misc{wu2026asleval,
  title     = {{ASLEval: Measuring Privacy Exposure Displacement in LLM Agent Sessions}},
  author    = {Wu, Guosen and Huang, Huizhen and Long, Guoxiong and Huang, Tao and Hou, Chen},
  howpublished = {arXiv (Cornell University)},
  year      = {2026},
  doi       = {10.48550/arxiv.2609.18864},
  url       = {https://doi.org/10.48550/arxiv.2609.18864}
}

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

Topic and keywords

Physical Sciences › Computer Science › Artificial Intelligence › Privacy-Preserving Technologies in Data

  • LLM agents
  • tool-using agents
  • privacy analysis

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?

Wu, G., Huang, H., Long, G., Huang, T., & Hou, C. (2026). ASLEval: Measuring Privacy Exposure Displacement in LLM Agent Sessions. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2609.18864

How many times has this paper been cited?

0 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?

Guosen Wu, Huizhen Huang, Guoxiong Long, Tao Huang, Chen Hou.

What is the DOI of this paper?

10.48550/arxiv.2609.18864, permanently at https://doi.org/10.48550/arxiv.2609.18864. 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.