Title: A Multi-Agent LLM Defense Pipeline Against Prompt Injection Attacks Show Chinese title

Title: 一种针对提示注入攻击的多智能体LLM防御流程

Abstract: Prompt injection attacks represent a major vulnerability in Large Language Model (LLM) deployments, where malicious instructions embedded in user inputs can override system prompts and induce unintended behaviors. This paper presents a novel multi-agent defense framework that employs specialized LLM agents in coordinated pipelines to detect and neutralize prompt injection attacks in real-time. We evaluate our approach using two distinct architectures: a sequential chain-of-agents pipeline and a hierarchical coordinator-based system. Our comprehensive evaluation on 55 unique prompt injection attacks, grouped into 8 categories and totaling 400 attack instances across two LLM platforms (ChatGLM and Llama2), demonstrates significant security improvements. Without defense mechanisms, baseline Attack Success Rates (ASR) reached 30% for ChatGLM and 20% for Llama2. Our multi-agent pipeline achieved 100% mitigation, reducing ASR to 0% across all tested scenarios. The framework demonstrates robustness across multiple attack categories including direct overrides, code execution attempts, data exfiltration, and obfuscation techniques, while maintaining system functionality for legitimate queries. Show Chinese abstract

Abstract: 提示注入攻击是大型语言模型（LLM）部署中的主要漏洞，其中嵌入在用户输入中的恶意指令可以覆盖系统提示并引发意外行为。本文提出了一种新颖的多智能体防御框架，该框架采用协调流水线中的专用LLM智能体来实时检测和中和提示注入攻击。我们使用两种不同的架构对我们的方法进行了评估：一种是顺序的智能体链流水线，另一种是基于分层协调器的系统。我们在55种独特的提示注入攻击上进行了全面评估，这些攻击分为8个类别，总共在两个LLM平台（ChatGLM和Llama2）上产生了400个攻击实例，结果表明安全性能显著提升。在没有防御机制的情况下，基线攻击成功率（ASR）达到了30% for ChatGLM和20% for Llama2。我们的多智能体流水线实现了100%的缓解，将ASR降低到所有测试场景中的0%。该框架在多个攻击类别中表现出鲁棒性，包括直接覆盖、代码执行尝试、数据泄露和混淆技术，同时保持了对合法查询的系统功能。