TechFlow reports, on September 3, OpenAI CEO Sam Altman stated that the Trump administration's voluntary review of the company's upcoming Astra model was a "productive process," and added that as AI capabilities continue to advance, strengthening communication with government officials worldwide will become increasingly important. OpenAI's upcoming Astra is the company's first model assessed with a "critical" cybersecurity capability rating.
OpenAI stated that models of this nature require more rigorous safety measures during both development and prior to public release. The Trump administration has established a voluntary framework for reviewing select advanced AI models before launch, which may include granting government officials up to 30 days of model access.
Altman confirmed that the Trump administration has already conducted a review of Astra, noting that the process is voluntary. "As these models reach significantly higher capability levels, I believe it will become increasingly important to fully implement this effort through close collaboration with security agencies in the US, UK, and other regions worldwide," Altman said, adding that Astra is an "important step forward."





