OpenAI has deployed a centralized control system called Control Room that can halt model output within 35 seconds of an intervention decision, according to a post on X. The system tracks all research inference workloads across OpenAI's infrastructure, including agent sampling for model training and evaluation, as well as data processing tasks. Control Room operates alongside OpenAI's misalignment monitoring system and is designed to quickly terminate or pause relevant tasks when safety hazards are detected. The company describes the system as part of its broader approach to safely training and deploying frontier models.