OpenAI launched its advanced LLM, Sol, for public access, a model considered on par with Anthropic's Fable, which was previously banned from public use due to government concerns. The approval process for such frontier models remains opaque, with former White House advisors and industry experts expressing confusion over licensing requirements and evaluation standards. Six cabinet agencies are instructed to determine a final process by early August, but no agreement exists on which models require scrutiny or which agencies should perform evaluations.
The lack of clear, standardized regulatory processes for frontier AI models creates an unpredictable operating environment for labs, favoring those with political connections.