Abstract
As Large Language Models (LLMs) become vital productivity tools, their integration into workflows poses significant risks without adequate safeguards. This paper proposes an open-source, privacy-focused, user-facing firewall designed to secure both web-based and programmatic LLM interactions.
Architecture
The architecture combines a browser extension and a proxy for total traffic interception across HTTP(S) and WebSocket communications. At its core, a flexible multi-agent pipeline delivers data leakage prevention through a hybrid approach that combines deterministic detectors with LLM-driven semantic analysis. This hybrid method not only prevents proprietary code leakage but also features extensible components for future security enhancements, such as prompt injection evasion.
Layered Framework
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.
Blogger's Review: This research not only highlights how to protect user privacy when interacting with large language models but also provides a scalable solution for future security measures. With the increasing prevalence of LLMs, this firewall architecture is particularly significant and warrants attention for its practical applications.