Today, Wednesday September 23, France -- which holds the UN Security Council presidency for September -- convened a 15-member session specifically on AI and international security. Sam Altman is confirmed to attend in person; Dario Amodei is expected to address the Council too, though by remote link rather than in the room. China's DeepSeek and fellow Chinese lab Moonshot were invited to make statements, though DeepSeek founder Liang Wenfeng is not attending in person. The session's stated focus: risks around loss of control, and the use of advanced AI models in actions affecting international peace and security. It's the first time the Council has directly hosted competing frontier Chinese and US developers together on shared safety concerns.
The meeting lands squarely inside a public fight over whether anyone should actually listen to Amodei's call to slow down. On September 14 at the All-In Summit in Los Angeles, Nvidia CEO Jensen Huang took a live on-stage phone call from President Trump, who dismissed AI-catastrophe warnings outright: "I'm telling you, it's all a hoax... the robots will not be taking over." Huang's on-stage reply, to applause: "We're not going to let that happen, sir."
The split has hardened into two named camps rather than a vague disagreement. Amodei, Altman, Elon Musk, and the ex-Anthropic researcher whose viral post kicked off this month's debate are on record warning of catastrophic risk; Trump, Vance, Zuckerberg and Huang are on record publicly rejecting a coordinated industry pause, with Zuckerberg specifically arguing that market incentives plus independent evaluators are sufficient safeguards without a coordinated slowdown. Separately, Zuckerberg, Musk and Huang reportedly lobbied Trump directly to block a proposed industry-funded AI regulator -- and Trump has since backed away from that proposal.
What's actually new today isn't the argument -- it's the venue. Up to now this has been a US domestic political fight playing out on stages and op-eds. A UN Security Council session is a different kind of room: its job is international security, not tech policy, and China's leading lab is sitting in it alongside the US labs making the opposite argument to their own government. I'd watch for one specific thing in the aftermath: does this session produce anything with actual teeth, or does it turn out to be the domestic fight restaged for an international audience with no more binding outcome than the essays already published this month.
The AI-safety-vs-acceleration split has moved from a US domestic fight to the UN Security Council, with competing US and Chinese labs both in the room for the first time -- the thing to track isn't who said what today, it's whether the session produces anything binding, since every prior round of this debate has ended in essays and phone calls, not commitments.