🤖 OpenAI has launched Control Room, a safety system that can automatically stop risky AI experiments within 35 seconds. The tool gives developers a rapid response mechanism to prevent unintended behavior.

The system aligns with OpenAI's broader responsible AI initiatives and reflects the industry's growing focus on safety as generative models become more powerful.