# DePrompt: Desensitization and Evaluation of Personal Identifiable Information in Large Language Model Prompts

arXiv: https://arxiv.org/abs/2408.08930 · HTML: https://mlsanity.com/paper/2408.08930/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 16 Aug 2024.

## Built on by (most cited first, 9 of 9)

- [Selective KV-Cache Sharing to Mitigate Timing Side-Channels in LLM Inference](https://mlsanity.com/paper/2508.08438/) (2025)
- [PII-Bench: Evaluating Query-Aware Privacy Protection Systems](https://mlsanity.com/paper/2502.18545/) (2025)
- [Anti-adversarial Learning: Desensitizing Prompts for Large Language Models](https://mlsanity.com/paper/2505.01273/) (2025)
- [PrivScope: Task-scoped Disclosure Control for Hybrid Agentic Systems](https://mlsanity.com/paper/2605.16630/) (2026)
- [Combating Knowledge Corruption in Agent Systems: A Byzantine-Tolerant Secure Collaborative RAG Framework](https://mlsanity.com/paper/2608.04366/) (2026)
- [What to Remember, What to Reveal: Privacy-Aware Memory for Conversational Agents](https://mlsanity.com/paper/2608.16551/) (2026)
- [One Size Fits All? A Modular Adaptive Sanitization Kit (MASK) for Customizable Privacy-Preserving Phone Scam Detection](https://mlsanity.com/paper/2510.18493/) (2025)
- [RedactionBench](https://mlsanity.com/paper/2606.18782/) (2026)
- [Privacy Personalization Trade offs in LLMs: The Impact of Stylometric Signal Reduction on User-Specific Text Generation](https://mlsanity.com/paper/2609.22112/) (2026)

## Abstract (arXiv)

Prompt serves as a crucial link in interacting with large language models (LLMs), widely impacting the accuracy and interpretability of model outputs. However, acquiring accurate and high-quality responses necessitates precise prompts, which inevitably pose significant risks of personal identifiable information (PII) leakage. Therefore, this paper proposes DePrompt, a desensitization protection and effectiveness evaluation framework for prompt, enabling users to safely and transparently utilize LLMs. Specifically, by leveraging large model fine-tuning techniques as the underlying privacy protection method, we integrate contextual attributes to define privacy types, achieving high-precision PII entity identification. Additionally, through the analysis of key features in prompt desensitization scenarios, we devise adversarial generative desensitization methods that retain important semantic content while disrupting the link between identifiers and privacy attributes. Furthermore, we present utility evaluation metrics for prompt to better gauge and balance privacy and usability. Our framework is adaptable to prompts and can be extended to text usability-dependent scenarios. Through comparison with benchmarks and other model methods, experimental evaluations demonstrate that our desensitized prompt exhibit superior privacy protection utility and model inference results.