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

“Exactly what that dialog looked like between the government and Anthropic and OpenAI is unclear.”