The AI Tactics Behind the Latest Cyber Threats

ReliaQuest's report finds that AI has become pervasive, commercialized, and weaponized in cybercrime—enabling LLM-generated malicious scripts, jailbroken models and malicious GPTs, deepfake-as-a-service, and autonomous exploitation frameworks that accelerate vulnerability discovery and exploitation (citing 45% of recent initial access attributed to vulnerability exploitation); notable examples include Skynet (prompt-injection evasion) and the Rhadamanthys infostealer, and the report urges defenders to prioritize logging, threat hunting, defense-in-depth, rapid remediation, and tighter AI governance.