OpenAI CEO Sam Altman was scheduled to brief the United Nations Security Council in person Wednesday on safety risks of increasingly capable AI, with Anthropic’s Dario Amodei joining by video alongside researchers including Yoshua Bengio, The Next Web and Security Council Report summaries said.

OpenAI separately published a proposal calling for the United States to lead international standards covering recursive self-improvement—systems that help build their own successors—days before the Xi state visit and after weekend talk of a U.S.–China AI dialogue.

Secretary-General António Guterres has warned about AI dangers at UNGA, while Trump dismissed global AI governance framing as a “globalist scheme” in his assembly speech. Accounts agree the Council session is a diplomatic first focused on loss-of-control risks; they diverge on whether U.S.-led standards can bind competitors or mainly shape allied procurement rules.

Officials and independent outlets continue updating confirmed details as they become available. NewsHabit’s card sticks to multi-outlet confirmation for this edition and notes where accounts still diverge, without projecting outcomes that the published record does not yet support.

Independent outlets continue to update confirmed details as officials brief and documents surface. NewsHabit keeps this card anchored to multi-outlet confirmation for the September 23, 2026 edition, flagging where accounts still diverge rather than projecting outcomes the published record does not yet support. Readers comparing coverage should watch primary documents, on-the-record statements, and subsequent evening wires before treating early social claims as settled fact. Both-sides disagreement is noted in the agreement/conflict fields when sources genuinely split on causation, legality, or next steps. Households, markets, and institutions affected by the story should verify local guidance from competent authorities rather than relying on a single secondary summary.