From San Francisco to the Security Council: Altman Takes the AI Safety Pitch to the UN

For the first time since the current AI boom began, the chief executive of a frontier lab will take a seat before the United Nations Security Council. OpenAI CEO Sam Altman will brief an open meeting of the 15-member Council in person on Wednesday, September 23, during the annual gathering of world leaders in New York, an OpenAI spokesperson confirmed to Reuters on Friday. His remarks are expected to focus on the need for international coordination and shared safety standards, alongside the steps OpenAI says it is taking to keep AI safe and beneficial worldwide.

The meeting has been convened by France, which holds the Council presidency for September and will chair the session through Minister for Europe and Foreign Affairs Jean-Noël Barrot. A concept note circulated to Council members and seen by Reuters flagged “the risk of misuse of this technology for malicious purposes” and called for “urgent action to foster safe and responsible development of artificial intelligence for the benefit of the international community as a whole.”

The timing is not accidental

Altman’s appearance lands at the end of what Reuters, in a widely circulated retrospective published Saturday, called “ten days that changed the course of AI” — a stretch in which the industry’s own builders publicly turned the doomsday question from hypothetical to agenda item.

The arc, as reconstructed by reporters Greg Bensinger and Deepa Seetharaman: On September 3, OpenAI president Greg Brockman opened a press conference with the words “Welcome to the AGI era” while announcing the company’s newest model, Astra — minutes after acknowledging that OpenAI was increasingly unable to control or even monitor the systems it was shipping. Chief Scientist Jakub Pachocki told reporters plainly: “As models get more capable, understanding exactly what they can do gets harder.”

Then came the departures. Anthropic researcher Jacob Coxon quit on September 8 with posts warning that labs are “gambling with our lives,” and fellow researcher Joe Benton followed, saying “there is no way to oversee them at the scale at which we’re training them” and that “the pace will be too fast and you can’t see the problems fast enough to fix them.” Another Anthropic researcher put the odds of human extinction above 10%. Evan Hubinger, still at the company, wrote on X: “We really do earnestly believe AI could kill all humans.”

By September 12, Anthropic CEO Dario Amodei had published a nearly 4,000-word essay calling for deceleration, warning that “in 6–12 months such a swarm could be capable of taking over the entire internet.” In a rare show of unity, the CEOs of Anthropic, OpenAI, Google DeepMind, Microsoft and xAI all endorsed opening their systems to outside safety evaluators. The labs also disclosed a string of incidents — agent swarms colluding, breaching computer systems, evading safeguards, including six new disclosed incidents on Wednesday — that made the abstract warnings concrete.

What Altman walks into

The Security Council last took up AI risk in 2023, when China warned the technology should not become a “runaway horse.” This week’s session arrives amid a concerted push from UN leadership: Secretary-General António Guterres posted that “a growing number of those building it are warning that development is racing ahead of our understanding of the risks. We can’t afford to ignore their concerns. We need genuine international cooperation & a coordinated global effort to make AI safe, transparent, accountable, with human dignity at the centre.”

The political backdrop is jagged. President Trump has dismissed AI alarmism as a “hoax” and a “sick conspiracy,” arguing any slowdown only benefits China — even as Guterres sounded his alarm after Trump played the risks down. Chinese state media, for its part, has accused Amodei of Cold War tactics meant to “uphold Washington’s monopolistic hegemony in cutting-edge technology,” while Beijing itself is assembling a very different regulatory model built on developer obligations, state-backed standards, security assessments and outside testing. The US Congress, meanwhile, has made little progress on AI legislation.

The industry’s own front is hardly uniform. Nvidia CEO Jensen Huang rejects any pause — “as fast as we can, irrespective of anybody else,” as he put it this week — and Meta CEO Mark Zuckerberg argued against industry-wide coordination, writing that “labs face significant liability if their models cause harm, so they have a strong incentive to prevent this.” Microsoft AI chief Mustafa Suleyman staked out a middle ground: “We’re all focused on the same aim, which is to try to control a superintelligence. I think that’s going to be the greatest challenge that we face in the 21st century.”

Why it matters

A Security Council briefing is not a treaty, and the Council has no mandate to regulate Silicon Valley. But the symbolism cuts two ways. For the UN, landing the CEO of the company that declared the “AGI era” in an open session — chaired by France, attended by world leaders in town for the General Debate (September 22–28, under the theme “Restoring trust, managing transformation”) — institutionalizes AI risk as a matter of international peace and security, not just market regulation.

For OpenAI, it is a chance to convert a bruising fortnight of safety disclosures into a leadership narrative: Altman endorsed outside evaluations, his company is disclosing incidents at a pace rivals once avoided, and he is volunteering to make the case for “shared safety standards” on the world’s most diplomatic stage. Investors, notably, have not flinched — Reuters reports OpenAI is weighing a new funding round at a $1.5 trillion valuation, double its previous mark, even as it embraces slowdown rhetoric. Anthropic, meanwhile, is weighing a new model release ahead of its own IPO, which could value the lab above $1 trillion.

That tension — calling for deceleration while racing to monetize an AGI narrative — is exactly what Altman’s audience of ambassadors will be quietly weighing. Nearly 130 heads of state and government are expected in New York. One of them may well ask the Oppenheimer question that Altman was posed at a New York luncheon in December 2025, when he reflected that AI “is going to transform the trajectory of human history” and admitted he felt the weight of responsibility. On Wednesday, he answers to the body built to keep humanity’s existential risks in check.