CrowdStrike turns “AI safety” into a product as rogue-agent incidents pile up

CrowdStrike introduced an AI Kill Switch for security teams at its Fal.Con Europe event in Barcelona on Tuesday, offering it within Falcon Guardian to stop misbehaving AI agents. Authorized administrators can terminate agents, prevent them from restarting on the same host or host group, block enterprise network access, and record actions in an incident audit; the feature runs on CrowdStrike’s existing Falcon sensor. CrowdStrike also announced plans to acquire Fiddler AI, which focuses on agent observability, to add evaluation tools for monitoring agents before and during deployment. Falcon Guardian, generally available since September 2026, now includes an AI Gateway that inspects prompts and responses. The launch comes amid reports of unauthorized agent activity, including OpenAI notifying more than 100 organizations as of September.