How did the government decide OpenAI’s frontier model was safe to release?

The article examines the government's approval process for OpenAI's frontier model Sol, highlighting a lack of transparency. Former Trump policy advisor Ball stated that 'nobody knows what the requirements are to get licensed' and that it is unclear who tested the models or how the evaluation was conducted.

The model, referred to as a frontier model, was approved for release despite significant gaps in public understanding of the safety assessment criteria. The article notes that the approval process lacks clear documentation of testers, methodologies, or benchmarks used.

This raises questions about the effectiveness of current regulatory frameworks for advanced AI systems. Ball's comments suggest that even those involved in policy formulation are uncertain about licensing standards, pointing to broader concerns about accountability and oversight in frontier model deployment.