“Ortermann: OpenAI Astra model was involved in voluntary government reviews in the US”

Ortman: OpenAI's Astra model participated in the U.S. government's voluntary review

OpenAI CEO Sam Altman said that the voluntary review of its upcoming Astra model by the Trump administration was a “productive process.” Astra will be the first model of the company to be classified as having “critical” cybersecurity capabilities. The Trump administration has established a voluntary framework for reviewing some advanced AI models before their release. Government officials may have up to 30 days to access the models, but the White House does not intend to make the framework public. Altman noted that as AI capabilities improve, close cooperation with AI security research institutions in the United States, the United Kingdom, and other parts of the world will become increasingly important. He also emphasized that people should be at the core of economic, social, and global decision-making, and that in the future, decisions should be made by humans rather than AI. Additionally, Jensen Huang warned at the G20 Innovation Ministers’ Meeting not to hinder AI development due to fear, while Altman called for a “realistic” and “accurate” approach to AI, avoiding doomism or blind optimism.