Why Tech's Biggest Names Are Walking Into the UN Security Council
This week was a very special one. The world's most powerful artificial intelligence builders met with the world's most powerful people involved in global security — and they acknowledged that these two worlds can no longer be separated.
Sam Altman of OpenAI, Dario Amodei of Anthropic and Hugging Face co-founder Clément Delangue spoke before the UN Security Council on Wednesday. The scene: the annual assembly of the UN General Assembly of world leaders. Whether artificial intelligence is progressing to a point where it could self-improve, evade any rational control and pose a real threat to international security.
This isn't a hypothetical panel. This marks the second occasion since 2023 that the Security Council has discussed AI head-to-head with China's concerns about AI as a "runaway horse" and the US' allegations of its use for censorship and repression.
The Ask: Benchmarks, Not Bans
Altman's warning entry was not alarmist, but rather practical. Altman intended to "make them think about common metrics" — tangible measures of what an AI can do and whether its protections are working — for world leaders, before handing over a briefing to reporters, said one OpenAI executive, who spoke to them in advance.
That's a different tone to the tune. Rather than calling on governments to halt AI advancement, the call is to create a shared scoreboard that is trustworthy.
The session was chaired by French Foreign Minister Jean-Noël Barrot. The top two protagonists in the development of AI, the United States and China, addressed council members directly. The Canadian researcher, Yoshua Bengio, who is sometimes referred to as one of the “Godfathers of AI”, also gave a briefing to the council, which is also co-chair of the UN's Independent International Scientific Panel.
The Real Fear: Algorithms Talking to Algorithms
A European diplomat, speaking on condition of anonymity, made it very clear. It's not just a rogue AI. A situation where automation – military or financial or otherwise – talks to automation and causes conflict without any human action involved.
But this is where the Security Council should be involved and not a tech regulator because of that risk. It's not just a market for artificial intelligence products but a matter of international security.
This is another time thing. The briefing was ahead of President Trump's meeting with Chinese leader Xi Jinping in Washington. The two countries have reportedly negotiated a communications mechanism for incidents involving AI that pose a national security threat — a minor but significant positive sign that the countries share a common risk.
Trump's Bet: Growth Over Guardrails
All the UN in this week do not feel that sense of urgency. Trump told the General Assembly he was "amazed" by AI and "will take action as a nation to do it carefully, but not to build a globalist plan to control it.
His message ran to the extreme: the United States is ahead of China and everyone else, and it means to remain that way. Instead of limiting the development of AI, Trump said the US would watch it by the Department of Justice.
That's a conflict with UN chief António Guterres, who has been warning for years that the world is facing AI that could pose grave risks – and who established an advisory committee of tech executives, officials, and academics last year to research just that.
Two Visions, One Table
The real tension that came out of the meeting on Wednesday was the conflict between two instincts: quick build and deal with risk as it arises, or slow down and set rules before the game begins. That balance has not been struck by AI leaders nor by the world's governments. This time, however, it is on the same stage, in the same room: at the UN Security Council, where the terms of the discussion are not "market share" but "international security".
