CrowdStrike has introduced a new ‘AI kill switch’ that allows businesses to shut down rogue artificial intelligence agents, prevent their restart, and cut off network access, according to CEO George Kurtz. This technology, part of the Falcon Guardian platform, addresses growing concerns about AI safety and is seen by Kurtz as a major growth opportunity for the cybersecurity firm.
The company highlighted strong customer demand, with enterprises requiring CrowdStrike’s integration before deploying AI offerings, demonstrating the critical need for robust AI safety measures in the current technological landscape.
READ MORE FROM CNBC
CrowdStrike CEO George Kurtz has announced the launch of a groundbreaking new technology designed to combat the growing risks associated with artificial intelligence. The company's new 'AI kill switch' provides businesses with the ability to stop rogue AI agents in their tracks, prevent them from reactivating, and sever their network access.
This innovative feature, unveiled at CrowdStrike's Fal.Con Europe conference, aims to bolster AI safety by giving human operators greater control over AI models and systems. "Everyone is building AI agents, but they never had the off button, and now they do," Kurtz stated in an interview with CNBC's Jim Cramer on "Mad Money." He emphasized its importance for ensuring the safe operation of AI.
The AI Kill Switch is integrated into CrowdStrike's Falcon Guardian platform, which is responsible for monitoring AI agents operating within an organization. Available immediately, this technology is a direct response to the increasing concerns surrounding the rapid deployment of AI agents. The announcement comes amidst broader discussions about AI safety, with figures like Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman calling for a slowdown in AI development due to safety risks. Kurtz, however, has previously argued that such a slowdown wouldn't eliminate existing risks and underscored the need for robust safeguards.
Kurtz views the ability to enable safe agent operation as a significant market opportunity for CrowdStrike. He highlighted the company's existing AI capabilities to prevent external breaches and their new capacity to ensure AI agents adhere to their intended instructions. He likened the technology to airbags in vehicles, stressing the necessity of safety mechanisms even for AI systems designed for correct operation. "If they go outside of those guardrails at runtime, we can kill those agents," he asserted.
The demand for CrowdStrike's security tools is reportedly surging as businesses increasingly adopt AI. Kurtz pointed to CrowdStrike's collaboration with Anthropic to secure its Claude Cowork product, noting that enterprise clients insisted on CrowdStrike's technology integration before deploying Anthropic's offerings. "Our customers went to Anthropic and said, 'We need CrowdStrike before we can deploy AI,'" Kurtz remarked, indicating that clients are unwilling to roll out AI without CrowdStrike's security assurances.
CrowdStrike's stock, which reached a record high of $278.86 on October 6, has experienced a decline of approximately 12% in line with broader tech stock struggles in October and November.
