App. Res. of Comp. Accepted Paper
Vol. 43, 2026 No. 7

Privacy protection via large language model unlearning: methods, evaluation, and challenges

Zhang Lei1,2,3
Zhang Qiang1,2,3
Qiao Junzhao1,2,3
Wu Mingxi1,2,3
Zhang Ning1,2,3
1. School of Information &Electronic Technology, Jiamusi University, Jiamusi Heilongjiang 154000, China
2. Heilongjiang Province Key Laboratory of Autonomous Intelligence & Information Processing, School of Information & Electronic Technology, Jiamusi University, Jiamusi Heilongjiang 154007, China
3. Jiamusi Key Laboratory of Satellite Navigation Technology & Equipment Engineering Technology, Jiamusi Heilongjiang 154007, China
DOI: 10.19734/j.issn.1001-3695.2025.12.0482

Abstract

The extensive deployment of large language models (LLMs) has heightened privacy risks such as membership inference and model inversion, making compliance with regulations including the right to be forgotten a key bottleneck in practice. Machine unlearning (MU) , which removes the influence of specific data after training, addresses this need for dynamic data deletion. This paper systematically surveyed MU methods for LLMs and proposed a novel taxonomy and evaluation framework. Existing methods were categorized into four classes tailored to LLM characteristics: data transformation, parameter fine-tuning, architecture design, and integrated training paradigms. The core mechanisms of these methods, covering reverse fine-tuning, parameter correction, architecture restructuring and privacy-aware training, were outlined. This paper constructed an evaluation framework based on unlearning completeness, utility preservation, and computational and communication overhead, enabling a comparative analysis of the strengths and applicability boundaries of each approach. Finally, within application scenarios such as federated learning and online services, unified formal definitions, verifiable unlearning, and support for streaming deletion remain critical open challenges.

Foundation Support

黑龙江省省属本科高校优秀青年教师基础研究支持计划(YQJH2024239)
黑龙江省自然科学基金联合基金培育项目(PL2024F002)
国家外国专家重点支撑项目(D20250185)
黑龙江省基本科研业务费基础研究项目(2019-KYYWF-1386)

Publish Information

DOI: 10.19734/j.issn.1001-3695.2025.12.0482
Publish at: Application Research of Computers Accepted Paper, Vol. 43, 2026 No. 7

Publish History

[2026-03-24] Accepted Paper

Cite This Article

张磊, 张强, 乔俊钊, 等. 基于大模型遗忘的隐私保护：方法、评估与挑战 [J]. 计算机应用研究, 2026, 43 (7). (2026-03-24). https://doi.org/10.19734/j.issn.1001-3695.2025.12.0482. (Zhang Lei, Zhang Qiang, Qiao Junzhao, et al. Privacy protection via large language model unlearning: methods, evaluation, and challenges [J]. Application Research of Computers, 2026, 43 (7). (2026-03-24). https://doi.org/10.19734/j.issn.1001-3695.2025.12.0482. )

About the Journal

  • Application Research of Computers Monthly Journal
  • Journal ID ISSN 1001-3695
    CN  51-1196/TP

Application Research of Computers, founded in 1984, is an academic journal of computing technology sponsored by Sichuan Institute of Computer Sciences under the Science and Technology Department of Sichuan Province.

Aiming at the urgently needed cutting-edge technology in this discipline, Application Research of Computers reflects the mainstream technology, hot technology and the latest development trend of computer application research at home and abroad in a timely manner. The main contents of the journal include high-level academic papers in this discipline, the latest scientific research results and major application results. The contents of the columns involve new theories of computer discipline, basic computer theory, algorithm theory research, algorithm design and analysis, blockchain technology, system software and software engineering technology, pattern recognition and artificial intelligence, architecture, advanced computing, parallel processing, database technology, computer network and communication technology, information security technology, computer image graphics and its latest hot application technology.

Application Research of Computers has many high-level readers and authors, and its readers are mainly senior and middle-level researchers and engineers engaged in the field of computer science, as well as teachers and students majoring in computer science and related majors in colleges and universities. Over the years, the total citation frequency and Web download rate of Application Research of Computers have been ranked among the top of similar academic journals in this discipline, and the academic papers published are highly popular among the readers for their novelty, academics, foresight, orientation and practicality.

Indexed & Evaluation

  • The Second National Periodical Award 100 Key Journals
  • Double Effect Journal of China Journal Formation
  • the Core Journal of China (Peking University 2023 Edition)
  • the Core Journal for Science
  • Chinese Science Citation Database (CSCD) Source Journals
  • RCCSE Chinese Core Academic Journals
  • Journal of China Computer Federation
  • 2020-2022 The World Journal Clout Index (WJCI) Report of Scientific and Technological Periodicals
  • Full-text Source Journal of China Science and Technology Periodicals Database
  • Source Journal of China Academic Journals Comprehensive Evaluation Database
  • Source Journals of China Academic Journals (CD-ROM Version), China Journal Network
  • 2017-2019 China Outstanding Academic Journals with International Influence (Natural Science and Engineering Technology)
  • Source Journal of Top Academic Papers (F5000) Program of China's Excellent Science and Technology Journals
  • Source Journal of China Engineering Technology Electronic Information Network and Electronic Technology Literature Database
  • Source Journal of British Science Digest (INSPEC)
  • Japan Science and Technology Agency (JST) Source Journal
  • Russian Journal of Abstracts (AJ, VINITI) Source Journals
  • Full-text Journal of EBSCO, USA
  • Cambridge Scientific Abstracts (Natural Sciences) (CSA(NS)) core journals
  • Poland Copernicus Index (IC)
  • Ulrichsweb (USA)