语义感知的 LLM 智能体以增强对话式 AI 服务中的隐私
Semantically-Aware LLM Agent to Enhance Privacy in Conversational AI Services
AI总结：
本文提出了LOPSIDED框架，通过动态替换用户提示中的敏感PII实体为语义一致的假名，在保护隐私的同时减少语义效用错误。
摘要：
随着对话式 AI 系统的日益普及，隐私泄露问题引发了越来越多的关注，特别是当用户在与大型语言模型（LLMs）交互时分享敏感个人数据。与这些模型共享的对话可能包含个人身份信息（PII），一旦暴露，可能导致安全漏洞或身份盗窃。为了应对这一挑战，我们提出了具有语义完整性的局部伪名化优化与定向实体检测（LOPSIDED）框架，这是一种语义感知的隐私智能体，旨在使用远程 LLM 时保护敏感 PII 数据。与通常会降低响应质量的先前工作不同，我们的方法动态地将用户提示中的敏感 PII 实体替换为语义一致的假名，从而保留对话的上下文完整性。一旦模型生成响应，假名会自动去伪名化，确保用户收到准确且保护隐私的输出。我们使用来自 ShareGPT 的真实对话来评估我们的方法，并对其进行了进一步的增强和标注，以评估命名实体是否与模型的响应在上下文上相关。我们的结果显示，与基线技术相比，LOPSIDED 将语义效用错误减少了 5 倍，同时增强了隐私。
英文摘要：
With the increasing use of conversational AI systems, there is growing concern over privacy leaks, especially when users share sensitive personal data in interactions with Large Language Models (LLMs). Conversations shared with these models may contain Personally Identifiable Information (PII), which, if exposed, could lead to security breaches or identity theft. To address this challenge, we present the Local Optimizations for Pseudonymization with Semantic Integrity Directed Entity Detection (LOPSIDED) framework, a semantically-aware privacy agent designed to safeguard sensitive PII data when using remote LLMs. Unlike prior work that often degrade response quality, our approach dynamically replaces sensitive PII entities in user prompts with semantically consistent pseudonyms, preserving the contextual integrity of conversations. Once the model generates its response, the pseudonyms are automatically depseudonymized, ensuring the user receives an accurate, privacy-preserving output. We evaluate our approach using real-world conversations sourced from ShareGPT, which we further augment and annotate to assess whether named entities are contextually relevant to the model's response. Our results show that LOPSIDED reduces semantic utility errors by a factor of 5 compared to baseline techniques, all while enhancing privacy.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
