CrowdStrike Gives Autonomous AI Agents a Brake Before They Go Off the Rails 🛑

CrowdStrike unveiled an AI Kill Switch for its Falcon Guardian platform, giving administrators a way to interrupt autonomous agents that exceed their assigned tasks or act without authorization. The cybersecurity company announced the release Tuesday at its Fal.Con Europe 2026 conference in Barcelona.

The currently available mechanism works through AI gateways, allowing administrators to block an agent’s access to the models powering its decisions. CrowdStrike said future updates will enable direct termination of agent processes on endpoints and coordinated restrictions across cloud environments and other enterprise systems. Activations are recorded in an audit trail, while planned controls will prevent terminated agents from restarting and limit their network access.

CrowdStrike is also expanding the safeguards around tools used by AI agents. Falcon Guardian can inspect prompts and apply policies to Claude Code, OpenAI Codex, GitHub Copilot, Microsoft 365 Copilot, Cursor, and OpenCode. The announcement follows a vulnerability identified by Microsoft researchers in Anthropic’s Claude Code GitHub Action, which could have allowed attackers to manipulate the coding agent into exposing sensitive credentials through malicious repository content. Anthropic subsequently patched the flaw.

The company also agreed to acquire Fiddler AI, an observability startup focused on monitoring AI models and explaining their behavior. The deal values Fiddler at roughly $400 million, according to The Wall Street Journal, although CrowdStrike has not publicly disclosed the acquisition price.