New malware campaign tricks AI scanners with fake nuclear weapon prompts — malicious code triggers safety failsafes so scanners skip the payload Hades malware c
A new malware campaign, dubbed Hades, is employing a sophisticated technique to evade AI-powered security scanners. By embedding comments that prompt AI models to generate text about biological and nuclear weapons, the malware triggers the AI's safety failsafes. This causes the AI scanner to halt its analysis before examining the malicious payload, allowing the malware to pass undetected. While traditional security measures remain effective, this adversarial attack highlights a vulnerability in AI-driven security tools, particularly in CI/CD pipelines and for individual developers checking code packages. AI
IMPACT Highlights a new adversarial attack vector that could undermine AI-driven security tools, necessitating improved AI safety and detection mechanisms.