Monday 16 November
Global

Your region, kept on this device.

OpenAI builds Control Room kill switch for its research models

Severity
Info
Sources
1

has deployed Control Room across its research infrastructure, a central system to track model workloads and terminate them when safety or security concerns arise. The stated stop latency is about 35 seconds, caused by a caching layer for inference authorisation, and OpenAI says it is working to reduce this to zero for instant job termination. The tool is designed to remain operational when other systems fail and to halt model output if Control Room itself goes down. No CVEs are involved, and OpenAI plans to expand the system alongside monitoring for misaligned behaviour and security breaches.

For security analysts · Written by AI from 1 source · Wrong?

What to do

  1. No action needed.

Written by AI from the sources, checked against them

Full story timeline · 1 update

  1. Wed 11 NovemberOpenAI builds a kill switch for its research modelsThe Register (Security)

Who reported · 1 outlet

The Register (Security)

Who and what is in it

Companies

Report a problem