Journal article · 2025 · Open access
Data Privacy Risks of Large Language Models: A Comprehensive Review
Applied and Computational Engineering, 218(1), 148–160 · Published 31 December 2025 · DOI 10.54254/2755-2721/2026.tj31027
In brief
“Data Privacy Risks of Large Language Models: A Comprehensive Review” is a 2025 journal article by Junkai Peng, published in Applied and Computational Engineering. OpenAlex counts 0 citations as of 19 November 2026. It is open access under a CC BY licence: the full text is free to read at Applied and Computational Engineering.
- Citations
- 0
- References
- 16
- Authors
- 1
Abstract
Large Language Models (LLMs) achieve strong performance across domains, yet persistent privacy risks remain. This paper characterizes privacy for LLMs as control over the collection, processing, sharing, and retention of sensitive information. Risks are grouped into data privacy and model privacy, then mapped to concrete attack surfaces. Data risks include training data leakage and membership inference. Model risks include model inversion, chain-of-thought leakage, Mixture-of-Experts routing leakage, and side-channel observation. Mitigation is organized at three technical levels and one governance layer. Data-level measures include cleaning, de-identification, and synthetic data. Model-level measures include differential privacy training, federated learning, chain-of-thought control, and backdoor defense. System-level measures include authenticated encryption, intrusion detection, traffic shaping, man-in-the-middle defense, and trusted hardware. Regulatory frameworks and standards offer reference points for deployment. Open problems include private inference for long contexts, quantitative privacy budgets in production telemetry, robust defenses for reasoning leakage, and integration with sparse MoE architectures. This perspective clarifies evaluation criteria and informs deployment choices under constraints.
Cite this paper
Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.
APA 7
Peng, J. (2025). Data Privacy Risks of Large Language Models: A Comprehensive Review. Applied and Computational Engineering, 218(1), 148–160. https://doi.org/10.54254/2755-2721/2026.tj31027
MLA 9
Peng, Junkai. “Data Privacy Risks of Large Language Models: A Comprehensive Review.” Applied and Computational Engineering, vol. 218, no. 1, 2025, pp. 148–160. https://doi.org/10.54254/2755-2721/2026.tj31027.
Chicago
Peng, Junkai. 2025. “Data Privacy Risks of Large Language Models: A Comprehensive Review.” Applied and Computational Engineering 218 (1): 148–160. https://doi.org/10.54254/2755-2721/2026.tj31027.
Harvard
Peng, J. (2025) ‘Data Privacy Risks of Large Language Models: A Comprehensive Review’, Applied and Computational Engineering, 218(1), pp. 148–160. doi:10.54254/2755-2721/2026.tj31027.
IEEE
J. Peng, “Data Privacy Risks of Large Language Models: A Comprehensive Review,” Applied and Computational Engineering, vol. 218, no. 1, pp. 148–160, 2025, doi: 10.54254/2755-2721/2026.tj31027.
BibTeX
@article{peng2025data,
title = {{Data Privacy Risks of Large Language Models: A Comprehensive Review}},
author = {Peng, Junkai},
journal = {Applied and Computational Engineering},
year = {2025},
volume = {218},
number = {1},
pages = {148--160},
doi = {10.54254/2755-2721/2026.tj31027},
url = {https://doi.org/10.54254/2755-2721/2026.tj31027}
}
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
- large language models
- data privacy
- model privacy
- training data leakage
- membership inference
- model inversion
- side-channel attacks
- differential privacy
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?
Peng, J. (2025). Data Privacy Risks of Large Language Models: A Comprehensive Review. Applied and Computational Engineering, 218(1), 148–160. https://doi.org/10.54254/2755-2721/2026.tj31027
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 Applied and Computational Engineering: sign in and open the PDF from this page.
Who wrote this paper?
Junkai Peng.
What is the DOI of this paper?
10.54254/2755-2721/2026.tj31027, permanently at https://doi.org/10.54254/2755-2721/2026.tj31027. 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.