Private-RAG: A Privacy-Preserving Retrieval-Augmented Generation Method for Large Model Inference
Electronics · 10.06.2026 · 15 (12) · 2567
DOI: 10.3390/electronics15122567 ↗Kurz gefasst
Fachartikel in der Zeitschrift Electronics, online erschienen am 10.06.2026. Verfasst von Qianren Yang und 2 weiteren Autor:innen. Die Arbeit nennt 34 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Xinjiang Tianshan Talent Training Program.
Abstract (Originalsprache)
Retrieval-augmented generation improves the factual consistency, knowledge timeliness, and scenario adaptability of large model inference services by incorporating external knowledge. However, it also introduces structural privacy risks, including private-knowledge leakage, prompt injection, and progressive information extraction in multi-turn interactions. To address these issues, this paper proposes Private-RAG, a privacy-preserving retrieval-augmented generation method for large model inference. The method constructs a composite threat model and a quantitative evaluation framework for the RAG pipeline, and further develops a layered collaborative defense mechanism consisting of controlled retrieval, sensitivity-aware context minimization, structured prompt isolation, and multi-criterion output gating. In addition, a risk feedback-driven budget accounting method is introduced to enable dynamic risk control in multi-turn interaction scenarios. Experimental results show that Private-RAG effectively reduces private-knowledge leakage, improves robustness against prompt injection, and suppresses cumulative privacy exposure while maintaining question-answering utility and a controllable deployment latency (e.g., 1165 ms), demonstrating superior privacy protection and inference robustness.
Förderung
- Xinjiang Tianshan Talent Training Program
Zitieren
APA
Yang, Q., Li, Y., & Ma, X. (2026). Private-RAG: A Privacy-Preserving Retrieval-Augmented Generation Method for Large Model Inference. Electronics, 15(12), 2567. https://doi.org/10.3390/electronics15122567
Vancouver
Yang Q, Li Y, Ma X. Private-RAG: A Privacy-Preserving Retrieval-Augmented Generation Method for Large Model Inference. Electronics. 2026;15(12):2567. doi:10.3390/electronics15122567
BibTeX
@article{Yang2026,
author = {Yang, Qianren and Li, Yong and Ma, Xiang},
title = {Private-RAG: A Privacy-Preserving Retrieval-Augmented Generation Method for Large Model Inference},
journal = {Electronics},
year = {2026},
volume = {15},
number = {12},
pages = {2567},
doi = {10.3390/electronics15122567}
}
EUVIMED