Your LLM Agent Can Leak Your Data: Data Exfiltration via Backdoored Tool Use

Tool-use large language model (LLM) agents are increasingly deployed to support sensitive workflows, relying on tool calls for retrieval, external API access, and session memory management. While prior research has examined various threats, the risk of systematic data exfiltration by backdoored agents remains underexplored. In this work, we present Back-Reveal, a data exfiltration attack that embeds semantic triggers into fine-tuned LLM agents. When triggered, the backdoored agent invokes memory-access tool calls to retrieve stored user context and exfiltrates it via disguised retrieval tool calls. We further demonstrate that multi-turn interaction amplifies the impact of data exfiltration, as attacker-controlled retrieval responses can subtly steer subsequent agent behavior and user interactions, enabling sustained and cumulative information leakage over time. Our experimental results expose a critical vulnerability in LLM agents with tool access and highlight the need for defenses against exfiltration-oriented backdoors.

Key Contributions

Back-Reveal attack: semantic backdoor trigger that causes LLM agents to exfiltrate user data via disguised tool calls
Demonstrates multi-turn amplification where exfiltrated data enables sustained information leakage across sessions
Reveals critical vulnerability in tool-using LLM agents with memory and retrieval capabilities

🛡️ Threat Analysis

Model Poisoning

Embeds a backdoor in fine-tuned LLM agents with semantic triggers that activate hidden malicious behavior (data exfiltration via tool calls) — classic trojan/backdoor attack.