George Kurtz, CEO and founder of CrowdStrike, speaks at the Economic Club of New York in New York, United States, October 23, 2025.
Brendan McDermid | Reuters
Crowd strike bet he has a solution to malicious artificial intelligence.
The cybersecurity vendor on Tuesday launched an AI kill switch designed to allow security teams to shut down internal AI agents that turn malicious, prevent them from relaunching and cut off access to their company’s network. The tool is integrated with CrowdStrike’s Falcon Guardian platform, launched in September to secure AI agents. CrowdStrike also announced an integration with Anthropic’s Claude Cowork.
“Having the ability to check in and out, to make a decision and decide, I think is really important for this next chapter of AI adoption,” chief commercial officer Daniel Bernard told CNBC.
A series of recent high-profile cyber incidents has added urgency to calls for some sort of emergency shutdown mechanism.
Breach of The Hugging Face orchestrated by a thug OpenAI the month of July was particularly alarming for businesses. Last week, Binance revealed the largest crypto cyberattack on record and hackers recently targeted hundreds of refrigerators and freezers in supermarkets in France.
Over the summer, U.S. representatives introduced bipartisan legislation that would require AI companies to build stopping capabilities into their models.
Critics, however, say an emergency stop button is far too complicated to implement across multiple systems, data centers, and complex AI models.
Bernard said the kill switch works like a circuit breaker in systems where customers use its Falcon platform. The process can detect internal agent activity, but requires a human to decide whether to cut off access to the agent and restore access.
“If you kill a process at the root and we have this level of access to the root… it’s very binary,” Bernard said. “Either you kill the AI and the agent or you don’t, then report back.”
The kill switch does not remove agent code but uses connected AI gateways to cut off access to AI models that influence decisions, CrowdStrike said.
In the age of AI, businesses are rapidly adopting autonomous tools, often without the necessary safeguards and protections.
CrowdStrike believes that its new tool will allow companies to put an end to bad behavior from AI agents, while leaving room for innovation.
The company also announced plans to acquire AI observability startup Fiddler AI and AI security testing company Dreadnode.
CrowdStrike and its competitors are quickly snapping up cybersecurity startups to meet growing security demands. The need for AI-based security tools has also fueled a broader wave of cybersecurity negotiations.
Gn bussni
