CrowdStrike CEO George Kurtz unveiled a new AI “kill switch” designed to immediately shut down rogue or malfunctioning AI agents, prevent them from restarting and disconnect them from corporate networks. Announced at the company’s Fal.Con Europe conference, the technology is part of its Falcon Guardian platform, which monitors AI agents operating within organizations and is available immediately. Kurtz said the tool addresses growing security concerns as businesses rapidly adopt autonomous AI systems, giving companies greater human control over their operations. He described AI agent safety as a major growth opportunity for CrowdStrike, comparing the safeguards to airbags in vehicles: even well-designed systems need protection when they operate outside established limits. The company is also benefiting from growing enterprise demand through its expanded partnership with Anthropic to secure its Claude Cowork platform. $CRWD