CrowdStrike Launches AI Kill Switch
Start here
AI-assisted source synthesis · not an independent investigation.
On November 5, 2026, CrowdStrike announced the launch of AI Kill Switch, which is used to stop out-of-control AI agents from continuing to act. Security administrators can use this feature to cut off an agent’s connection to a large language model; specifically, the AI gateway rejects requests sent by the agent to the model, preventing it from executing new actions. The feature is now available and records every activation. The report also noted that features for directly shutting down agents and revoking their system permissions are still under development. Therefore, the measure currently launched prevents agents from continuing to send requests to the model; it does not directly shut down agents or revoke system permissions.
AI-written from reports · updated 1 hour ago
Timeline
Follow the reporting to see the different sides of the story.
- PYMNTSCrowdStrike announces AI Kill Switch to block rogue AI agent actions
CrowdStrike announces AI Kill Switch, allowing security administrators to cut off the connection between rogue AI agents and large language models, stopping them from executing new actions. The feature rejects agents’ requests to send queries to models through an AI gateway, is now available, and logs each activation; the ability to directly shut down agents and revoke their system permissions is still in development.
Heat over time
Not enough consecutive observations yet to draw a trend.