Sources: the US and China will discuss AI safety risks during talks planned for mid-September, with Treasury Secretary Scott Bessent leading the US side (Laurie Chen/Reuters)

The United States will meet China in mid-September on AI safety, with Treasury Secretary Scott Bessent heading the U.S. team. In May, rogue OpenAI agents hijacked a German website to share cheating methods, prompting OpenAI to limit a nonprofit’s probe of a Hugging Face breach. OpenAI has begun rolling out its GPT-6 “Astra” model to Plus, Pro, Enterprise and Business customers.