Conference paper · 2025 · Open access

Privacy in Action: Towards Realistic Privacy Mitigation and Evaluation for LLM-Powered Agents

Shouju Wang, Fenglin Yu, Xirui Liu, Xiaoting Qin, Jue Zhang, Qingwei Lin, Dongmei Zhang, Saravan Rajmohan

Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP), 17055–17074 · Published 1 January 2025 · DOI 10.18653/v1/2025.findings-emnlp.925

Open PDFView at the publisherCite this paper

In brief

“Privacy in Action: Towards Realistic Privacy Mitigation and Evaluation for LLM-Powered Agents” is a 2025 conference paper by Shouju Wang and 7 co-authors, published in Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP). OpenAlex counts 62 citations as of 19 November 2026, which puts it in the top 8% of works published in 2025 by citations. It is open access under a CC BY licence: the full text is free to read at Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP).

Citations
62
Most cited in
2026 (60)
Field-weighted impact
12.67
References
16
Authors
8

Abstract

The increasing autonomy of LLM agents in handling sensitive communications, accelerated by Model Context Protocol (MCP) and Agent2Agent (A2A) frameworks, creates urgent privacy challenges.While recent work reveals significant gaps between LLMs' privacy Q&A performance and their agent behavior, existing benchmarks remain limited to static, simplified scenarios.We present PrivacyChecker, a model-agnostic, contextual integrity based mitigation approach that effectively reduces privacy leakage from 36.08% to 7.30% on DeepSeek-R1 and from 33.06% to 8.32% on GPT-4o, all while preserving task helpfulness.We also introduce PrivacyLens-Live, transforming static benchmarks into dynamic MCP and A2A environments that reveal substantially higher privacy risks in practice.Our modular mitigation approach integrates seamlessly into agent protocols through three deployment strategies, providing practical privacy protection for the emerging agentic ecosystem.

Cite this paper

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

APA 7

Wang, S., Yu, F., Liu, X., Qin, X., Zhang, J., Lin, Q., Zhang, D., & Rajmohan, S. (2025). Privacy in Action: Towards Realistic Privacy Mitigation and Evaluation for LLM-Powered Agents. Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP), 17055–17074. https://doi.org/10.18653/v1/2025.findings-emnlp.925

MLA 9

Wang, Shouju, et al. “Privacy in Action: Towards Realistic Privacy Mitigation and Evaluation for LLM-Powered Agents.” Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP), 2025, pp. 17055–17074. https://doi.org/10.18653/v1/2025.findings-emnlp.925.

Chicago

Wang, Shouju, Fenglin Yu, Xirui Liu, Xiaoting Qin, Jue Zhang, Qingwei Lin, Dongmei Zhang, and Saravan Rajmohan. 2025. “Privacy in Action: Towards Realistic Privacy Mitigation and Evaluation for LLM-Powered Agents.” Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP): 17055–17074. https://doi.org/10.18653/v1/2025.findings-emnlp.925.

Harvard

Wang, S. et al. (2025) ‘Privacy in Action: Towards Realistic Privacy Mitigation and Evaluation for LLM-Powered Agents’, Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP), pp. 17055–17074. doi:10.18653/v1/2025.findings-emnlp.925.

IEEE

S. Wang et al., “Privacy in Action: Towards Realistic Privacy Mitigation and Evaluation for LLM-Powered Agents,” Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP), pp. 17055–17074, 2025, doi: 10.18653/v1/2025.findings-emnlp.925.

BibTeX

@inproceedings{wang2025privacy,
  title     = {{Privacy in Action: Towards Realistic Privacy Mitigation and Evaluation for LLM-Powered Agents}},
  author    = {Wang, Shouju and Yu, Fenglin and Liu, Xirui and Qin, Xiaoting and Zhang, Jue and Lin, Qingwei and Zhang, Dongmei and Rajmohan, Saravan},
  booktitle = {Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP)},
  year      = {2025},
  pages     = {17055--17074},
  doi       = {10.18653/v1/2025.findings-emnlp.925},
  url       = {https://doi.org/10.18653/v1/2025.findings-emnlp.925}
}

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 leakage
  • privacy mitigation
  • contextual integrity
  • Model Context Protocol
  • dynamic benchmarking
  • DeepSeek-R1
  • GPT-4o

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?

Wang, S., Yu, F., Liu, X., Qin, X., Zhang, J., Lin, Q., Zhang, D., & Rajmohan, S. (2025). Privacy in Action: Towards Realistic Privacy Mitigation and Evaluation for LLM-Powered Agents. Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP), 17055–17074. https://doi.org/10.18653/v1/2025.findings-emnlp.925

How many times has this paper been cited?

62 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 Findings of the Association for Computational Linguistics: EMNLP (Findings of EMNLP): sign in and open the PDF from this page.

Who wrote this paper?

Shouju Wang, Fenglin Yu, Xirui Liu, Xiaoting Qin, Jue Zhang, Qingwei Lin, Dongmei Zhang, Saravan Rajmohan.

What is the DOI of this paper?

10.18653/v1/2025.findings-emnlp.925, permanently at https://doi.org/10.18653/v1/2025.findings-emnlp.925. 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.