It is not every week that the people racing to build a technology stand in front of the world’s most powerful security body and ask to be restrained. That is roughly what happened at the United Nations Security Council on Wednesday, 23 September 2026, during the 81st General Assembly.
“If managed poorly, I even believe AI could be a risk to humanity as a whole,” Dario Amodei, chief executive of Anthropic, told the 15-member council. Sam Altman, head of OpenAI, echoed him, saying humanity could “lose control of the future of AI.”
Altman made the sharper argument for why it matters where the rules come from: “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 they serve.”
Why the UN
The meeting was convened by France, and the foreign ministers of France and the United Kingdom said the international community needed to act. The council is the body built for problems that cross borders and threaten everyone, which is how the industry’s own leaders now describe their work. It is a remarkable shift in framing: AI is no longer a product story. It is being argued as a security story.
The contradiction
Here is the part that should be held up to the light. Two of the loudest warnings came from executives whose companies are in a race to deploy the same systems faster than anyone else. A call for regulation from the front of the pack is not the same as a call to slow down. It can also be a call to write the rules that rivals will have to follow.
And the governments with the most leverage are not lining up behind a pause. Reporting this week notes that both the United States and China have shunned calls to slow the technology, even as AI leaders warn of catastrophe. In the US, lawmakers have moved in the opposite direction at home, proposing sweeping restrictions including a ban framed around “superintelligence,” while the administration has pushed its diplomats toward different language entirely.
The honest read
The gap between what AI leaders say at the UN and what their companies and governments do next is the real story. Watch the specifics: whether any binding framework follows, who writes it, and whether the firms asking for rules would accept ones that actually constrain them. Warm words at a Security Council meeting cost nothing. A signed limit costs everything.
Sources
- Al Jazeera and agencies, “OpenAI, Anthropic CEOs call for global AI regulation at UN,” 24 September 2026 — https://www.aljazeera.com/news/2026/9/24/ai-corporate-leaders-tell-un-the-industry-needs-global-regulation
- Al Jazeera, “As AI leaders warn of catastrophe, US and China shun slowdown calls,” 23 September 2026; and “US lawmakers propose sweeping AI restrictions with superintelligence ban,” 23 September 2026
Leave a Reply