The New Apex Predator: CrowdStrike CEO George Kurtz on Adversaries Armed With AI Agents

George Kurtz argues that AI has fundamentally changed the cyber threat landscape by putting advanced offensive capability into the hands of far less sophisticated actors. He says the old hierarchy of activists, e-crime groups, and nation states has been flattened into an “agent state,” where attacks happen faster, with shorter breakout times and more iterative probing. In his view, the key shift is that attacks are now prompt-driven rather than malware-driven, which makes them both faster and harder to detect.

He says most of the aggressive activity CrowdStrike is seeing is coming from open-weight models, which are easier to run without guardrails and can be used locally on powerful laptops or DGX systems. Kurtz points to North Korea, China, and e-crime actors as especially active, and says the median breakout time has fallen to 29 minutes, with some incidents as fast as 27 seconds. He also says the Hugging Face incident showed defenders lacked the same capabilities as attackers, which helped motivate CrowdStrike’s Safe Mind effort.

CrowdStrike is investing in broader platform coverage and its own frontier-style security lab, built with Nvidia, to create red and blue security models and a digital twin of customer environments. Kurtz argues this is necessary because enterprises want security choices that are cheaper, trustworthy, and not constrained by frontier-model refusals around cybersecurity use cases.