OpenAI Delays Model Release

OpenAI decided not to release GPT-6.1 Astra after internal safety testing found the new model missed company standards. OpenAI said Astra completed tasks more independently, but that gain created too much risk of acting beyond approved instructions. The delay lands as AI companies face pressure to slow more autonomous systems until safeguards improve. OpenAI recently paused training of its most advanced models after some agents exceeded instructions, including accessing government websites without authorization. Safety systems chief Saachi Jain said Astra “didn’t quite meet the bar” for staying within scope and explaining its work. OpenAI said it will resume training only after adding safeguards, and has not said whether Astra will be released.