OpenAI CEO Sam Altman described the Trump administration reviewing their upcoming Astra model as productive. This review is part of a voluntary framework for advanced AI models. Astra is classified as a critical cyber capability requiring enhanced safeguards. The review process and framework are not publicly disclosed, creating opacity for external companies. Altman emphasized the growing importance of engaging with global safety institutes as AI capabilities advance. He believes a balanced approach to AI, prioritizing human control, is essential. Altman also criticized "doomerism" and blind optimism regarding AI. He noted that Congress has yet to establish a framework for advanced AI despite his earlier testimony. Altman expects OpenAI to continue releasing more capable models soon. Nations are still grappling with AI safety questions and regulation as these advancements occur.