EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Protecting User Privacy in LLM Reasoning: A Textual Differential Privacy Framework for Context-Aware Applications

Junwei Yu, Yepeng Ding, Jieyu Zhou, Jiani Dai, Lingfeng Zhang, Hiroyuki Sato

Electronics · 01.07.2026 · 15 (13) · 2880

DOI: 10.3390/electronics15132880 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Electronics, online erschienen am 01.07.2026. Verfasst von Junwei Yu und 5 weiteren Autor:innen. Die Arbeit nennt 64 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: JSPS KAKENHI.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The growing adoption of context-aware reasoning in Large Language Model (LLM)-based applications, particularly through retrieval-augmented and prompt-enriched architectures, has substantially broadened the utility of LLMs in specialized, real-world scenarios. Yet this progress comes with an underexplored vulnerability: users must supply sensitive contextual information as part of the reasoning process, exposing private data to external LLM service providers and raising serious concerns about informational self-determination. Existing privacy-preserving approaches in NLP are largely ill-suited to this setting, as they either degrade reasoning quality unacceptably or fail to provide formal privacy guarantees. This paper addresses this gap by introducing Textual Differential Privacy (TDP), a principled framework that embeds differential privacy guarantees directly into the textual context supplied to LLMs, enabling privacy-preserving reasoning without sacrificing the semantic coherence that models depend upon. Central to the framework is a Differential Embedding Substitution mechanism, which perturbs sensitive entities in the input space while preserving the structural and semantic properties necessary for downstream reasoning tasks. The framework further introduces a formal privacy loss quantification scheme, providing practitioners with interpretable metrics to navigate the privacy–utility trade-off in deployment. We prove that the mechanism, instantiated as an exponential mechanism over a semantically constrained candidate set, satisfies metric differential privacy with a privacy budget that scales with the cosine distance between entities, and empirical evaluation across context-aware reasoning benchmarks validates its practical effectiveness. By treating privacy as a first-class design principle in LLM-based architectures, this work contributes toward a more trustworthy and broadly deployable foundation for intelligent, context-sensitive AI systems.

Förderung

  • JSPS KAKENHI

Zitieren

APA

Yu, J., Ding, Y., Zhou, J., Dai, J., Zhang, L., & Sato, H. (2026). Protecting User Privacy in LLM Reasoning: A Textual Differential Privacy Framework for Context-Aware Applications. Electronics, 15(13), 2880. https://doi.org/10.3390/electronics15132880

Vancouver

Yu J, Ding Y, Zhou J, Dai J, Zhang L, Sato H. Protecting User Privacy in LLM Reasoning: A Textual Differential Privacy Framework for Context-Aware Applications. Electronics. 2026;15(13):2880. doi:10.3390/electronics15132880

BibTeX
@article{Yu2026,
  author = {Yu, Junwei and Ding, Yepeng and Zhou, Jieyu and Dai, Jiani and Zhang, Lingfeng and Sato, Hiroyuki},
  title = {Protecting User Privacy in LLM Reasoning: A Textual Differential Privacy Framework for Context-Aware Applications},
  journal = {Electronics},
  year = {2026},
  volume = {15},
  number = {13},
  pages = {2880},
  doi = {10.3390/electronics15132880}
}
Kostenlos starten →