ICON: Indirect Prompt Injection Defense for Agents based on Inference-Time Correction
This addresses security vulnerabilities in LLM agents for users deploying them in real-world applications, offering a balanced defense against attacks while maintaining workflow efficiency.
The paper tackles the problem of Indirect Prompt Injection attacks on LLM agents by proposing ICON, a framework that detects attacks via latent space signatures and corrects them with attention steering, achieving a 0.4% attack success rate and over 50% task utility gain.
Large Language Model (LLM) agents are susceptible to Indirect Prompt Injection (IPI) attacks, where malicious instructions in retrieved content hijack the agent's execution. Existing defenses typically rely on strict filtering or refusal mechanisms, which suffer from a critical limitation: over-refusal, prematurely terminating valid agentic workflows. We propose ICON, a probing-to-mitigation framework that neutralizes attacks while preserving task continuity. Our key insight is that IPI attacks leave distinct over-focusing signatures in the latent space. We introduce a Latent Space Trace Prober to detect attacks based on high intensity scores. Subsequently, a Mitigating Rectifier performs surgical attention steering that selectively manipulate adversarial query key dependencies while amplifying task relevant elements to restore the LLM's functional trajectory. Extensive evaluations on multiple backbones show that ICON achieves a competitive 0.4% ASR, matching commercial grade detectors, while yielding a over 50% task utility gain. Furthermore, ICON demonstrates robust Out of Distribution(OOD) generalization and extends effectively to multi-modal agents, establishing a superior balance between security and efficiency.