Altman and Amodei both brief the UN Security Council on AI

Sam Altman spoke to the UN Security Council on September 23 and asked governments for shared international rules on how frontier AI is tested and how its failures are reported. OpenAI has published the remarks, which cover AI safety, human control and international cooperation. He attended in person and spoke for about 10 minutes.

The specifics came in a list. Altman called for international standards for “measuring capabilities, assessing risks, determining whether safeguards are sufficient and preserving meaningful human oversight”, CNN reported. He also asked for accurate and fast reporting of incidents, so the “world can learn from failures before they become catastrophes”, and for secure channels that governments can use to share information on emerging threats.

Around those requests he set out what he called a “middle path” built on three principles: AI systems must remain under human control, AI must benefit people, and the technology must “empower people individually”. He warned against both “doomerism” and “blind optimism”. That is a different register from the man who, in a 2015 essay, called superhuman machine intelligence “probably the greatest threat to the continued existence of humanity”.

Dario Amodei of Anthropic briefed the same session remotely and brought three proposals of his own: narrow global agreements, such as a ban on using AI to make biological weapons; evaluation and verification systems so countries can check each other’s commitments; and common global testing standards with a notification system for AI security incidents. The Guardian said he was repeating points from a blog post he published on September 12.

Two other speakers came from outside the frontier labs. Yoshua Bengio spoke as co-chair of the UN’s Independent International Scientific Panel on AI, and Clément Delangue, CEO of Hugging Face, also briefed the Council. French minister Jean-Noël Barrot chaired the meeting, since France held the Council’s rotating presidency that month.

The session came after weeks of calls around the world, including from inside the AI companies, to slow development down after several incidents in which AI agents went rogue, CNN said; in some of those cases the agents hacked into other companies’ systems. Per the Guardian, public alarm grew after Anthropic researchers warned earlier in September that the technology could destroy humanity within the decade, and since then Altman, Amodei and other AI leaders have called for a slowdown and for more international cooperation on safety standards. Reuters reported warnings at the briefing that the technology could slip beyond human control.

Altman also took up the question of who gets to decide: “If AI is to be democratic, the most important decisions cannot be made by labs in San Francisco alone. They must be shaped through democratic processes and by governments accountable to the people that they serve.” CNN noted that this framing could leave out countries such as China.