arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2607.08282cs.CRcs.AIcs.MA
收藏

用于在与语言模型交互中保护敏感数据隐私的多智能体防火墙架构

Multi-Agent Firewall Architecture for Privacy Protection of Sensitive Data in Interactions with Language Models

Universidad Carlos III de Madrid(卡洛斯三世大学)

Hugo García Cuesta, Pablo Mateo Torrejón, Alfonso Sánchez-Macián

AI总结:

针对大语言模型集成到工作流程中缺乏保障措施的风险,提出开源多智能体防火墙架构,结合浏览器扩展和代理拦截流量,通过混合方法防止数据泄漏,分层架构可跨异构环境部署,最佳配置下F1分数达94.93%。

摘要:

虽然大语言模型(LLMs)已成为重要的生产力工具,但在没有充分保障措施的情况下将其集成到工作流程中会带来重大风险。本文提出了一个面向用户的、专注于隐私的开源防火墙,用于保护基于网络和编程的LLM交互安全。该架构结合了浏览器扩展和代理,以拦截HTTP(S)和WebSocket通信的所有流量。其核心是一个灵活的多智能体管道,通过将确定性检测器与LLM驱动的语义分析、专有代码泄漏预防以及为未来安全增强(如提示注入规避)设计的可扩展组件相结合的混合方法来防止数据泄漏。该框架的分层架构允许在异构环境中部署,使组织能够平衡计算成本、检测深度和延迟。评估结果表明,在最佳配置下,它的F1分数高达94.93%。

英文摘要:

While Large Language Models (LLMs) have become essential productivity tools, their integration into workflows without adequate safeguards creates significant risks. This paper proposes an open-source, privacy-focused, user-facing firewall designed to secure both web-based and programmatic LLM interactions. The architecture combines a browser extension and a proxy for total traffic interception across both HTTP(S) and WebSocket communications. At its core, a flexible multi-agent pipeline delivers data leakage prevention through a hybrid approach combining deterministic detectors with LLM-driven semantic analysis, proprietary code leakage prevention, and extensible components designed for future security enhancements such as prompt injection evasion. The framework's layered architecture enables deployment across heterogeneous environments, allowing organizations to balance computational cost, detection depth and latency. Evaluation results demonstrate it achieves F1 scores of up to 94.93% on optimal configurations.

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。

↑