Safeguarding System Prompts for LLMs

Abstract: LLMs are increasingly utilized in applications where system prompts, which guide model outputs, play a crucial role. These prompts often contain business logic and sensitive information, making their protection essential. However, adversarial and even regular user queries can exploit LLM vulnerabilities to expose these hidden prompts. To address this issue, we propose PromptKeeper, a robust defense mechanism designed to safeguard system prompts. PromptKeeper tackles two core challenges: reliably detecting prompt leakage and mitigating side-channel vulnerabilities when leakage occurs. By framing detection as a hypothesis-testing problem, PromptKeeper effectively identifies both explicit and subtle leakage. Upon detection, it regenerates responses using a dummy prompt, ensuring that outputs remain indistinguishable from typical interactions when no leakage is present. PromptKeeper ensures robust protection against prompt extraction attacks via either adversarial or regular queries, while preserving conversational capability and runtime efficiency during benign user interactions.