ClawGuard: Out-of-Band Detection of LLM Agent Workflow Hijacking via EM Side Channel
Leo Linqian Gan, Jeffery Wu, Longyuan Ge, Lanqing Yang, Yonghao Song + 4 more
TLDR
ClawGuard uses electromagnetic side channels to out-of-band detect workflow hijacking in LLM agents, offering a forge-resistant security solution.
Key contributions
- Detects LLM agent workflow hijacking by monitoring EM emanations.
- Leverages unique hardware usage patterns of agent skills for detection.
- Uses external software-defined radios and a drift-aware pipeline for analysis.
- Achieves 100% true-positive rate and 1.16% false-positive rate on a large dataset.
Why it matters
This paper introduces a novel, out-of-band security defense for LLM agents, addressing a critical vulnerability where host-internal methods can be compromised. ClawGuard uses physical EM side channels to provide a robust, forge-resistant mechanism, significantly enhancing the trustworthiness and security of autonomous AI workflows.
Original Abstract
Autonomous LLM agents face a critical security risk known as workflow hijacking, where attackers subtly alter tool and skill invocations. Existing defenses rely on host-internal telemetry (such as audit logs), which can be forged if the host OS is compromised. To solve this, we introduce ClawGuard, a passive, out-of-band monitor that audits LLM-agent workflows using electromagnetic (EM) emanations. Because distinct agent skills create unique hardware usage patterns (computation, DRAM, network blocking), they emit measurable, macroscopic EM envelopes. External software-defined radios (SDRs) capture these physical signals. Using a drift-aware pipeline with 320-dimensional features, ClawGuard converts RF streams into physical evidence. Evaluated on a 7.82TB RF corpus, ClawGuard achieved an AUC of 0.9945, detecting attacks with a 100% true-positive rate and a 1.16% false-positive rate. This proves passive EM sensing is a practical, forge-resistant physical check against compromised host software.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.