OpenAI and Anthropic executives urged AI coordination at the UN after Trump rejected global oversight.
Why it matters: The fight over AI governance is moving into a live policy arena, with the UN, the White House and frontier AI labs all trying to shape the rules. That could affect how safety standards, capability testing and human control are handled worldwide.
- Sam Altman, Dario Amodei and Clément Delangue briefed the UN Security Council in New York on Sept. 23, 2026.
- Trump told the UN General Assembly on Sept. 22 that the U.S. “totally rejects” any “globalist scheme” to control AI.
- OpenAI has called for “compatible international approaches” to measuring capabilities, managing risk and preserving human control.
- Anthropic said AI could fully automate or sharply accelerate work relevant to international security as soon as early 2027.
Top executives from OpenAI, Anthropic and Hugging Face used a UN Security Council briefing in New York to push for international coordination on AI risks, according to the official transcript from the Sept. 23 meeting here.
The session came one day after President Donald Trump told the UN General Assembly that the United States “totally rejects” any attempt to create a “globalist scheme” to control artificial intelligence, a sharp rebuke to multilateral oversight efforts reported by Reuters.
OpenAI has separately argued that governments should adopt “compatible international approaches” to measuring capabilities, managing risk, preserving human control and deciding when development should slow or stop, in a Sept. 9 policy post published by the company.
Anthropic’s Frontier Safety Roadmap says the company believes AI could, as soon as early 2027, fully automate or dramatically accelerate work in domains relevant to international security and the global balance of power in its roadmap.
France, which held the Security Council presidency for the meeting, described AI as an urgent international security issue and said the world needs a balanced approach with humans at the center of decision-making in the record.
The Security Council has previously taken up AI as a peace-and-security issue, including meetings in 2023 and 2024, as part of a broader UN governance push documented in earlier transcripts.
By the numbers
- Sept. 23, 2026 - date of the Security Council briefing in New York
- Early 2027 - Anthropic's timeline for possible AI capability gains relevant to security
- 2023 and 2024 - prior UN Security Council discussions of AI as a security issue
Yes, but: It is not clear from the available sources whether the Security Council produced any formal outcome, declaration or statement from the session.