According to Axios, OpenAI CEO Sam Altman said the Trump administration's voluntary review of OpenAI's forthcoming Astra model was a "productive process" and confirmed the company participated in it. Altman said Astra is a "significant step forward" and noted that as AI systems become more capable, closer engagement with safety institutes in the U.S., the U.K. and elsewhere will matter more. He said OpenAI expects to keep releasing models as countries work through safety questions, and he urged a "realistic" and "accurate" approach to AI regulation. Altman also said he expected Congress to create a "basic framework" for advanced AI after his 2023 testimony, but that has not happened.