OpenAI Deploys Control Room to Halt Risky Research Workloads
OpenAI says it has deployed Control Room across its research infrastructure to track inference workloads, including agentic sampling for model training and evaluation, and data processing. Working with misalignment monitoring systems, it can stop or pause a job and halt model output within 35 seconds after an intervention decision. The company also says the system uses fail-closed behavior, denying inference requests if Control Room itself fails, is disabled, or is tampered with.
Control Room turns OpenAI’s model-behavior monitoring into an operational shutdown mechanism for internal research workloads, including agentic training tasks.
As our models get more powerful and the associated safety and security risks become bigger, it becomes more and more critical that we can detect potential issues, and quickly stop models in their tracks when needed. To solve this, we built Control Room.
Control Room is a central system to track model workloads across our research infrastructure and stop them quickly when safety or security concerns arise. It connects every research inference workload, including agentic sampling for model training and evaluation, data processing, and more. This works together with our misalignment monitoring systems, and halts model output within 35 seconds of a decision to intervene to stop or pause a job. It has been built to keep working even when other systems fail, and to stop model output if Control Room itself goes down.
Control Room is already deployed across our research infrastructure, where the need is greatest. Our external model-serving stack is already much more centralized by nature, with a well-developed operational capability to quickly take down model deployments if needed.
How it works
Control Room works by integrating a few systems: the workload, the inference engine, and the centralized Control Room service.