Control Room

OpenAI research infrastructure for registering agent jobs and revoking further inference output; workload code can continue running.

Records citing Control Room

Nov 9, 2026
OpenAI Control Room revokes further agent inference without terminating workload code
DefenseFrameworkOpenAI

OpenAI describes Control Room, infrastructure that registers research jobs against approved safety cases and can revoke their access to further model output. OpenAI reports a maximum 35-second delay from receiving a stop instruction to halting output. Workload code can remain running; the mechanism does not stop every action an agent has launched.