According to Jin10, OpenAI said its largest planned frontier reinforcement learning training remains paused while safety checks are completed.