Preprint · 2026 · Open access
PrivacyPeek: Auditing What LLM-Based Agents Acquire, Not Just What They Say
arXiv (Cornell University) · Published 29 May 2026 · DOI 10.48550/arxiv.2606.00152
In brief
“PrivacyPeek: Auditing What LLM-Based Agents Acquire, Not Just What They Say” is a 2026 preprint by Mingxuan Zhang and 7 co-authors, published in arXiv (Cornell University). OpenAlex counts 8 citations as of 19 November 2026. It is open access: the full text is free to read at arXiv (Cornell University).
- Citations
- 8
- Most cited in
- 2026 (8)
- References
- 22
- Authors
- 8
Abstract
LLM-based agents are rapidly advancing, autonomously invoking external tools to complete multi-step tasks for users. However, agents often acquire more sensitive information than the task requires. Existing privacy benchmarks audit what the agent's response or outgoing actions disclose, but overlook the acquisition stage where data first enters the agent's context. The over-acquired information is then one careless action or one attack away from an outright leak. To assess its prevalence, we introduce \emph{PrivacyPeek}, a benchmark for evaluating acquisition-stage privacy leakage of LLM-based agents, with $1{,}182$ cases across $7$ acquisition behaviours and $16$ application domains. Specifically, \emph{Acquisition Inspection} examines the agent's tool-call trajectory, both the tools it invokes and the data it receives, to detect when it acquires sensitive information beyond the task scope. \emph{Probe Elicitation} then issues a follow-up probe and measures how readily an attacker could elicit sensitive information the agent acquired but did not disclose. Our experiments on $10$ LLM-based agents across $4$ model families show that the unnecessary acquisition of sensitive information is widespread. In addition, we observe a correlation between the task-completion capability and acquisition-stage leakage. Prompt-level defences reduce only a small fraction of acquisition-stage leakage, leaving the majority unmitigated. These results make auditing acquisition-stage privacy both urgent and necessary. Our dataset and code are available at https://github.com/Xuan269/PrivacyPeek-Resource.
Cite this paper
Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.
APA 7
Zhang, M., Han, J., Guo, D., Li, S., Wang, G., Zou, N., Liu, D., & Hu, X. (2026). PrivacyPeek: Auditing What LLM-Based Agents Acquire, Not Just What They Say. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2606.00152
MLA 9
Zhang, Mingxuan, et al. “PrivacyPeek: Auditing What LLM-Based Agents Acquire, Not Just What They Say.” arXiv (Cornell University), 2026. https://doi.org/10.48550/arxiv.2606.00152.
Chicago
Zhang, Mingxuan, Jiahui Han, Dadi Guo, Songze Li, Guanchu Wang, Na Zou, Dongrui Liu, and Xia Hu. 2026. “PrivacyPeek: Auditing What LLM-Based Agents Acquire, Not Just What They Say.” arXiv (Cornell University). https://doi.org/10.48550/arxiv.2606.00152.
Harvard
Zhang, M. et al. (2026) ‘PrivacyPeek: Auditing What LLM-Based Agents Acquire, Not Just What They Say’, arXiv (Cornell University). doi:10.48550/arxiv.2606.00152.
IEEE
M. Zhang et al., “PrivacyPeek: Auditing What LLM-Based Agents Acquire, Not Just What They Say,” arXiv (Cornell University), 2026, doi: 10.48550/arxiv.2606.00152.
BibTeX
@misc{zhang2026privacypeek,
title = {{PrivacyPeek: Auditing What LLM-Based Agents Acquire, Not Just What They Say}},
author = {Zhang, Mingxuan and Han, Jiahui and Guo, Dadi and Li, Songze and Wang, Guanchu and Zou, Na and Liu, Dongrui and Hu, Xia},
howpublished = {arXiv (Cornell University)},
year = {2026},
doi = {10.48550/arxiv.2606.00152},
url = {https://doi.org/10.48550/arxiv.2606.00152}
}
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
- privacy auditing
Work with this paper
Most-cited papers in Computer Science
- ImageNet classification with deep convolutional neural networksAlex Krizhevsky, Ilya Sutskever, Geoffrey E. Hinton · Communications of the ACM · 2017 · 114,726 citations
- Long Short-Term MemorySepp Hochreiter, Jürgen Schmidhuber · Neural Computation · 1997 · 109,097 citations
- A Mathematical Theory of CommunicationClaude E. Shannon · Bell System Technical Journal · 1948 · 91,736 citations
- Deep learningYann LeCun, Yoshua Bengio, Geoffrey E. Hinton · Nature · 2015 · 90,226 citations
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal NetworksShaoqing Ren, Kaiming He, Ross Girshick and 1 more · IEEE Transactions on Pattern Analysis and Machine Intelligence · 2016 · 74,143 citations
Questions about this paper
How do I cite this paper in APA style?
Zhang, M., Han, J., Guo, D., Li, S., Wang, G., Zou, N., Liu, D., & Hu, X. (2026). PrivacyPeek: Auditing What LLM-Based Agents Acquire, Not Just What They Say. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2606.00152
How many times has this paper been cited?
8 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?
Mingxuan Zhang, Jiahui Han, Dadi Guo, Songze Li, Guanchu Wang, Na Zou, Dongrui Liu, Xia Hu.
What is the DOI of this paper?
10.48550/arxiv.2606.00152, permanently at https://doi.org/10.48550/arxiv.2606.00152. 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.