The headline. CrowdStrike opened Fal.Con Europe 2026 in Barcelona on November 2 with one message: companies will only let AI agents (software that takes actions on its own, not just answers questions) loose at scale if humans can stop them. Almost every announcement this week hangs off that idea.
The Kill Switch. The new AI Kill Switch in Falcon Guardian lets an authorized admin stop a rogue agent where it runs, block it from launching again on that machine or group of machines, and cut its access to the company network. Every action is logged. CrowdStrike says it is available today. One caveat in the fine print: termination works “first at the Gateway, and soon on the endpoint,” so the laptop-level stop is still coming.
What George Kurtz said. From CrowdStrike’s press release: “AI safety is how you get the full benefit of AI. When you deploy agents with a kill switch, you stay in control.” And: “For the first time, security is simultaneously the accelerant and the safety net for technology adoption. The AI Kill Switch puts power in the hands of the people to realize AI safety.” These are the quotes CrowdStrike published. We have not verified a transcript of his keynote remarks.
Why it matters. At Fal.Con in Las Vegas in September, theCUBE Research reported that CrowdStrike’s AJ Shipley said the company did not then have a generalized kill switch. Two months later, it ships one. That is fast product work, and it tells you buyers were asking.
The two acquisitions. CrowdStrike signed deals to buy Fiddler AI, which watches what agents do and explains why (it becomes Falcon Observe), and Dreadnode, which attacks AI models and agents with jailbreaks and prompt injection before they go live. Both are expected to close and ship in CrowdStrike’s fiscal Q4 FY27 (November to January). CrowdStrike disclosed no price, only that it will pay mostly cash plus some vesting stock. The Wall Street Journal reported about $400M for Fiddler and $95M for Dreadnode, roughly $495M combined. Treat those as reported, not confirmed.
The rest of the product news. Falcon AI Gateway is now generally available and included with Falcon Guardian. Guardian’s AI threat detection now runs on SafeMind, CrowdStrike’s own security models built with NVIDIA. Agent hooks cover Claude Code, OpenAI Codex, GitHub Copilot, OpenCode and Cursor. Falcon Complete for Guardian brings 24/7 managed detection and response to AI security. Guardian is also coming natively to Anthropic’s Claude Cowork, and Falcon Shield now covers apps built on Lovable.
The partners. Dynatrace is teaming with CrowdStrike to go after legacy SIEM (the big log-search systems SOCs run on) and observability. CoreWeave is bringing Falcon and 24/7 MDR to its AI cloud. Dell AI Factory with NVIDIA customers can choose CrowdStrike for AI security. Deloitte Spain expanded its partnership into AI risk advisory with Falcon IQ. Project QuiltWorks, CrowdStrike’s program for fixing flaws that frontier AI finds faster, added 15 European partners including Atos, BT, Computacenter, Deutsche Telekom Security, Orange Cyberdefense and Telefónica.
The takeaway. CrowdStrike is now selling the full agent lifecycle: test it (Dreadnode), watch it (Fiddler), stop it (Kill Switch). For practitioners, the question to put to any vendor is simple: who can stop an agent, how fast, and on which machines today. For analysts, roughly $495M (reported) is small against a company worth about $248B. The real bet is that AI security becomes a new line on the Falcon bill.