Sam Altman to brief UN Security Council on AI dangers

A rare Security Council moment for AI

Sam Altman, the chief executive of OpenAI, is scheduled to brief the United Nations Security Council in person next week during the annual gathering of world leaders for the UN General Assembly in New York, according to an OpenAI spokesperson cited by Reuters [1]. The 15-member Council is due to meet on Wednesday on artificial intelligence and international security, a framing that places frontier AI not only in the technology or economic policy category, but inside the UN body responsible for international peace and security [1].

The meeting is being convened by France, which holds the rotating presidency of the Security Council for September, and is expected to be chaired by French Foreign Minister Jean-Noël Barrot [1]. That choice of chair matters: it signals that Paris wants the discussion to be treated as a high-level diplomatic issue, not merely as a technical briefing. France’s concept note, seen by Reuters, says the Council should examine the risk that AI could be misused for malicious purposes and the need to foster safe and responsible development for the benefit of the international community as a whole [1].

Altman’s expected message is tightly aligned with that diplomatic framing. An OpenAI spokesperson said his remarks are expected to focus on international coordination, shared safety standards, and the steps OpenAI says it is taking to ensure that AI is safe and benefits people globally [1]. In other words, Altman is not simply appearing as the head of a powerful technology company; he is entering a forum where governments are likely to ask whether private AI labs should help shape global safeguards, be subject to them, or both.

Why the timing is politically charged

The Security Council briefing comes amid renewed public anxiety over how quickly the most capable AI systems are advancing. Reuters reported that the Council meeting follows calls from several AI industry leaders for a coordinated slowdown in the development of increasingly powerful technology, with warnings that such systems could soon improve on their own and slip beyond human control [1]. That fear — loss of control over rapidly improving models — appears to be one of the central themes France wants the Council to examine: the concept note says the meeting should allow participants to analyze risks linked to loss of control or to the use of the latest AI models in ways that could affect international peace and security [1].

The diplomatic backdrop is also crowded. CNN reported that the Trump administration plans to convene its own high-level AI event on the sidelines of the UN General Assembly on the same Wednesday as the Security Council meeting [2]. According to CNN, the US event is being pitched as a chance to discuss partnership with the United States on AI, including the policy, legal and regulatory conditions that would allow adoption at scale, with free-expression considerations expected to be part of the conversation [2]. That creates two parallel tracks: a Security Council session focused on AI risks and international security, and a US-hosted event emphasizing AI adoption, partnership and governance conditions [2].

The contrast is not subtle. CNN reported that President Donald Trump has pushed back against government regulation and described concerns about the technology as a “hoax” [2]. The Associated Press, in reporting carried by The Washington Post, also described Trump’s pushback, noting that he called threats that AI might destroy humanity a “HOAX” and argued that efforts to limit the technology were part of a “SICK conspiracy” [3]. That means Altman’s UN appearance is unfolding in a political environment where calls for shared safeguards are being weighed against fears that regulation could slow US innovation or advantage competitors.

Guterres puts the UN behind global guardrails

UN Secretary-General António Guterres sharpened the global-governance case on Wednesday, warning that the world cannot afford what he called a “race to the bottom on AI safety” [3]. He told UN reporters that national action is essential but global coordination is indispensable, and he called for guardrails that make AI safe, transparent, accountable and centered on human dignity [3]. His comments give the upcoming Council meeting a broader UN context: the issue is not merely whether OpenAI, Anthropic, Google DeepMind or other companies slow down, but whether states can agree on some minimum rules for systems whose effects cross borders.

Guterres also said AI will be a major topic in his discussions with world leaders during the high-level week of the General Assembly [3]. That is important because the Security Council is not the only UN venue grappling with AI. Still, it is the venue that frames issues through peace, security, conflict, sanctions, cyber threats and the behavior of major powers. If AI is becoming a Council issue, it is because policymakers increasingly see the technology as relevant to crisis management, disinformation, autonomous systems, cyberattacks and strategic competition.

The Associated Press reported that Guterres referred to a proposal by Anthropic chief executive Dario Amodei for a coordinated slowdown in AI development by leading technology companies and governments [3]. The same report said Altman, Elon Musk and other prominent industry figures had recently said AI development needs to slow down, while Meta chief executive Mark Zuckerberg distanced his company from that position and said every AI company has to ensure its technology is safe [3]. That split within the technology sector gives the Security Council debate an added layer: the industry is not speaking with a single voice, even as it asks governments to take its warnings seriously.

What Altman is likely to be asked

Altman’s appearance will likely put three questions before Council members. First, what kinds of AI risks are severe enough to be treated as international security issues? France’s concept note points to malicious misuse, loss of control, and the use of latest-generation models to develop, facilitate or conduct actions affecting peace and security [1]. That could include cyber operations, disinformation campaigns, biological-risk assistance, military decision support, or autonomous targeting — even if the public reporting ahead of the meeting has not specified a negotiated outcome.

Second, what would shared safety standards actually mean? An OpenAI spokesperson said Altman is expected to call for international coordination and common safety standards [1]. The phrase sounds straightforward, but it raises difficult questions: whether standards should be voluntary or legally binding; whether they should apply only to the most capable systems; who would evaluate models; whether audits would be public; and how countries would handle companies operating across jurisdictions.

Third, who gets a seat at the table? CNN quoted Claire Melamed of the United Nations Foundation as saying there is growing interest in a UN role in crisis management, especially if an AI-generated cyberattack spills across borders and affects institutions in countries with fewer governance resources [2]. She also said she does not expect a firm deliverable from the Security Council meeting because the permanent members of the Council have very different positions [2]. That warning is central: the United States, China, Russia, France and the United Kingdom all have strategic AI interests, and any serious Council process would have to navigate great-power rivalry.

The Council’s authority — and its limits

The Security Council’s involvement gives the discussion symbolic force, but it does not automatically produce rules. Reuters noted that the Council first met on AI risks in 2023, when China warned that the technology should not become a “runaway horse” and the United States warned against using AI to censor or repress people [1]. That earlier session showed that members can agree AI is consequential while disagreeing sharply on the greatest risks and the right remedies.

The same dynamic is visible now. Guterres is urging cooperation and common guardrails, but AP reporting highlights obstacles including domestic and global competition, especially from China, and profit incentives inside the AI race [3]. CNN’s reporting similarly suggests that the UN is still trying to define its role, with many AI-focused side events around the General Assembly but uncertainty over how central the topic will be in formal UN business [2].

That does not make the meeting irrelevant. On the contrary, the Council may be useful precisely because it forces governments to speak about AI in security terms before a crisis compels them to do so. A formal treaty is not expected from one open briefing. But the session can normalize the idea that AI safety is not only a corporate responsibility or a domestic regulatory question. It can also create pressure for more regular briefings, expert channels, model-evaluation norms, and coordination mechanisms for cross-border incidents.

Why this briefing matters

Altman’s scheduled appearance at the Security Council marks a new stage in the public politics of AI risk. The head of the company behind ChatGPT is not going to New York simply to explain a product; he is entering a chamber where the language is peace, security, misuse, control and collective responsibility [1]. That setting will make it harder for governments to treat frontier AI as a purely commercial race, and harder for companies to argue that voluntary internal safeguards are enough.

The meeting may also expose the gap between urgency and consensus. France is emphasizing malicious use and loss of control, Guterres is calling for global guardrails, and the United States is hosting a parallel event focused on AI partnership and adoption conditions [1][2][3]. Those tracks can complement one another, but they can also pull in different directions if “safety” is heard in one room as a global obligation and in another as a constraint on innovation.

The most realistic expectation is not a dramatic Security Council decision. It is a diplomatic marker. If Altman’s briefing leads Council members to discuss AI risks more systematically, it could help move the issue from occasional alarm to continuing governance. If it produces only speeches, the meeting will still show that the debate over frontier AI has reached the world’s most powerful security forum — and that the question is no longer whether AI belongs on the international security agenda, but what governments are prepared to do now that it is there.

Developments

1. OpenAI's Sam Altman to brief UNSC on AI Risks amid Rising ConcernsTRT World · Sep 18, 2026, 11:51 PM UTC · 8/10

Sources from the last 72 hours

1. [1]Exclusive-OpenAI's Sam Altman to brief UN Security Council next weekSep 18, 2026, 8:27 PM UTC
2. [2]Trump admin to host UN event on AI next weekSep 18, 2026, 10:16 PM UTC
3. [3]UN chief urges global AI cooperation saying world can't afford 'a race to the bottom on AI safety'Sep 16, 2026, 8:41 PM UTC

AI-generated article based on recent web research, then preserved as a dated editorial snapshot.