Briefs / Technology / U.S. rejects global AI governance as Altman, Amodei urge UN oversightTechnology briefing
U.S. rejects global AI governance as Altman, Amodei urge UN oversight
At a France-convened UN Security Council meeting on AI and international security, OpenAI’s Sam Altman and Anthropic’s Dario Amodei pressed for global oversight while White House OSTP Director Michael Kratsios said the U.S. “totally reject[s]” international bodies asserting centralized control of AI.
By US Brief desk · Updated 2026-09-24T15:57:00-07:00
Editor’s note: Facts and quotes follow the UN Security Council 10228th-meeting press coverage (Sep 23, 2026), the U.S. Mission to the UN as-delivered intervention by Michael Kratsios, and multi-outlet wraps from Al Jazeera (AP/Reuters), The Next Web, and related reporting dated Sep 23–24, 2026. Distinguish this governance-diplomacy story from the separate Australian Medicare portal breach (tech-openai-medicare). Do not invent Council resolutions—the meeting was a debate/briefing, not a binding vote. Carolina Principles and Xi-visit AI agenda are attributed as reported context, not as completed bilateral deals.
At a France-convened United Nations Security Council meeting on artificial intelligence and international security during UN General Assembly week, frontier-lab CEOs urged global oversight frameworks even as the Trump administration’s science adviser told the Council the United States rejects international bodies asserting centralized control of AI.
What happened: On September 23, the Council’s 10228th meeting heard briefers including OpenAI CEO Sam Altman, Anthropic CEO Dario Amodei, Hugging Face CEO Clément Delangue, and Yoshua Bengio, co-chair of the UN Independent International Scientific Panel on AI. Al Jazeera’s wrap with AP and Reuters, published September 24, and the UN’s own meetings coverage summarize a sharp split: industry and several European ministers pressed for common standards, while Washington drew a hard line against “global governance” of what U.S. officials increasingly call superintelligence.
What the CEOs said: Amodei told members that if managed poorly, AI “could be a risk to humanity as a whole,” and urged industry-wide, global-scale risk management—including narrow agreements that could win broad support, such as bans on using AI for biological weapons, plus shared evaluation and testing standards. Altman warned humanity could “lose control of the future of AI,” argued that “if AI is to be democratic, the most important decisions cannot be made by labs in San Francisco alone,” and said those choices “must be shaped through democratic processes and by governments accountable to the people they serve.” UN coverage also recorded his caution against both “blind optimism” and “doomerism.”
Delangue and Bengio: Delangue told the Council Hugging Face had publicly disclosed an autonomous-agent cyberattack earlier in 2026 and had “defended ourselves with AI,” including reliance on a Chinese model that faced fewer restrictions than comparable U.S. tools—an asymmetry point he used to argue for stronger monitoring, incident-disclosure standards, and open approaches to cybersecurity. Bengio warned that “the dangers are real and imminent,” telling the Council it faces a threat “none can contain alone” and that does not respect borders; The Next Web reported he went further in calling for frontier AI to be licensed.
Washington’s rejection: Michael Kratsios, assistant to the president and director of the White House Office of Science and Technology Policy, delivered the U.S. intervention. In the as-delivered text posted by the U.S. Mission to the UN, he said the frontier’s advance “is not a reason to pause its further development or to constrain it with new global governance structures,” and that international dialogue “cannot be allowed to drift toward global governance.” Echoing President Trump’s General Assembly remarks, Kratsios said the United States “totally rejects any attempt to construct a globalist scheme of control for superintelligence.” Al Jazeera’s paraphrase—“We totally reject all efforts by international bodies to assert centralised control and global governance of AI”—matches that line’s substance.
National sovereignty pitch: Kratsios pointed to recent G20 innovation-minister consensus on the Carolina Principles, which he said agree countries should “preserve national sovereignty in the governance of emerging technologies” rather than cede it to international bodies. He urged other states to legislate for their own people, build domestic adoption capacity, and use U.S. export partnerships for “sovereign” infrastructure—arguing the Council should share best practices, not create a global regulator. “A prosperous future will not be secured by a global regulator,” the as-delivered text concludes.
Why it matters: The meeting lands amid a U.S.–China technology race and an expected Xi Jinping visit to Washington that Al Jazeera and others say is likely to touch whether and how to regulate AI—even as both capitals remain reluctant to lock in major new restrictions that could slow their own labs. It also sits beside parallel AI-safety politics already on US Brief’s Technology board, including the separate Australian Medicare statistics-portal episode involving an OpenAI agent (tech-openai-medicare)—a product-security controversy, not this diplomacy fight.
Who is affected: Frontier labs whose CEOs are asking governments for rules while competing on capability; U.S. allies hearing a sovereignty-first export pitch; Global South members who, in the same debate, warned about equitable access and “digital colonialism”; and investors watching whether UN talk hardens into disclosure, licensing, or biosecurity norms—or stays rhetorical while national rules dominate.
Uncertainty and competing views: France’s foreign minister and the UK’s Ed Miliband argued governments cannot outsource protection to private companies and that common frameworks are needed; Russia questioned whether generic AI belongs on the Council’s agenda; China rejected “us-versus-them” tech blocs and highlighted its own governance initiatives. No binding Council resolution emerged from the debate in the sources reviewed. Whether “global oversight” means soft scientific panels and dialogue forums already created by the General Assembly, or harder licensing and verification regimes, remains contested—and U.S. policy as stated rejects the latter.
What to watch next: Any concrete follow-on from Finland/Norway-style common-framework calls; whether Trump–Xi talks produce even symbolic AI language; U.S. domestic moves that operationalize the Carolina Principles and export program; and industry incident-disclosure standards after agent-misuse cases. For readers, the confirmed story is a high-profile Security Council airing of AI risk in which major U.S. lab chiefs urged international guardrails while the White House publicly rejected centralized global governance of AI—not a new UN treaty.
Follow US Brief: Bluesky · X · RSS
Full interactive version requires JavaScript. https://us-brief.com/article/tech-un-ai-council
Skip to content
