According to TechFlow, on September 13, jin10.com reported that OpenAI founder Sam Altman publicly stated that he fully aligns with Anthropic CEO Dario's core viewpoint that "frontier model progress needs to be controlled." This issue has become a key focus of recent internal discussions at OpenAI.
Regarding AI safety governance, a topic of widespread concern across the industry, Altman noted that granting independent third-party evaluators system access equivalent to that of internal staff is a highly valuable initiative. He explicitly pledged that OpenAI will adopt this framework and implement it.
Specific regulatory guidelines and evaluation frameworks are currently under development. The relevant parties have indicated that more substantive progress and accompanying measures will be disclosed to the public in the near future.





