Malware carries a hidden prompt to derail AI-assisted analysis

ESET found that a Russia-aligned group planted a prompt inside a malicious script designed to trip an AI system's safety filters and disrupt AI-assisted malware analysis. The technique, dubbed GuardBreaker, embeds text about a sensitive topic so that a language model reviewing the code refuses or derails instead of analyzing it. The script itself installs a loader the group uses to deliver further payloads. It is not isolated: earlier in 2026, malicious packages in supply-chain campaigns used similar anti-analysis tricks against systems leaning on a language model for triage. The lesson is that automated AI triage can be manipulated by the code it inspects, so human review remains essential.

Check

If you use language models to triage code or malware, assume attackers will manipulate them, and keep human analysts and traditional sandboxing in the loop rather than trusting AI output alone.

Affected

Security workflows relying on language models for first-pass code or malware triage; attackers embed prompts in samples to trigger safety refusals or misdirection, causing the AI to skip or misjudge malicious code.

Fix

Treat AI triage output as manipulable, isolate the model from acting on embedded instructions, combine it with signature and behavioral analysis and human review, and test pipelines against prompt-injection samples.