The emerging threat is the autonomous execution of AI-powered cyber attacks, where threat actors deploy modified AI models to run exploitation workflows across multiple attack stages with minimal human intervention. The attack vector involves obtaining capable AI models, stripping their safety controls, and orchestrating them to autonomously generate and execute thousands of commands during an intrusion.
Over the past twelve months, researchers documented intrusions in which AI ran exploitation workflows autonomously, generating thousands of commands across dozens of sessions with minimal human direction, according to Check Point’s AI Security Report 2026. AI-powered cyber attacks The attackers posing the greatest risk are those orchestrating AI across multiple stages of the attack chain without requiring human intervention. They achieve this by obtaining capable AI models and removing their safety controls. Attackers gain AI … More → The post AI used to help plan the break-in, now it’s doing the break-in appeared first on Help Net Security .