OpenAI's Sasha Baker 'cautiously optimistic' on US-China AI talks

Tuesday's remarks arrived alongside competing signals about who should govern advanced AI. Baker framed a possible US-China notification system as "substantive and real," while also stressing it was "a starting place, not the ending place" — a deliberate hedge that leaves room for deeper cooperation without promising it. She tied the need for dialogue to the fact that both countries host the world's leading AI labs, but added that "every country in the world has a stake," suggesting the conversation should not stay bilateral.

The remarks also previewed a split-screen week: OpenAI's Sam Altman was set to bring the standards message to the UN Security Council on Wednesday, the same week President Donald Trump rejected international oversight of AI as a "globalist scheme." How AI governance gets decided — through bilateral channels, multilateral bodies, or national action — remains unsettled.

What happens next likely hinges on whether the reported notification system moves past discussion. If it does, it could give AI labs and their government-affairs teams a formal venue for safety coordination across the two countries; if it stalls, the starting point may not advance.