Book a Call
Things to Know

Free resources · Policy

Anthropic’s CEO told the UN AI is “the most important global security issue facing the world today”. Are you ready?

What Anthropic’s CEO asked the UN Security Council, and how to read it

A viral clip of Dario Amodei at the Security Council is real. Here is what he proposed, and the context that matters.

Andrea d’Agostini, Co-Founder and CEO of Monarchy Media3 min read

Questions this article answers

  1. What did Dario Amodei say at the UN Security Council?
  2. What does “a country of geniuses in a data center” mean?
  3. Is AI writing 90% of code, as Dario Amodei predicted?
  4. What is Anthropic’s Responsible Scaling Policy?
  5. Has the UN Security Council discussed AI before?

Should Ask Questions

SAQ: the questions you should be asking (and nobody is)

  1. Did you check that clip before you shared it?
  2. Does the person warning you about AI risk have a commercial stake in AI?
  3. Have the predictions of the people building AI come true?
WatchVideo summary

A clip of Anthropic’s chief executive, Dario Amodei, briefing the UN Security Council is circulating widely. Before using it, we checked it against the UN record. The remarks are genuine.

This article sets out what the Council discussed, what Amodei proposed, and the context a careful reader needs: his commercial interest, his track record on timelines and his record on oversight. It closes with what his proposals mean for any organisation that relies on AI. Our own opinion is labelled as Monarchy’s view.

From the video summary: Anthropic’s CEO told the UN AI is “the most important global security issue facing the world today”. Are you ready?
From the video summary

01What a frontier-lab CEO asked the Security Council

On 23 September 2026, at its 10228th meeting, the UN Security Council took up "Artificial intelligence and international security" under France's presidency. The briefers were Yoshua Bengio, co-chair of the UN's Independent International Scientific Panel on AI, and the chief executives of OpenAI, Anthropic and Hugging Face 12. A clip of Anthropic's Dario Amodei, who spoke by video link, is circulating. We checked it against the UN record before using it. The UN's transcript, the UN Web TV video of his 5-minute briefing and C-SPAN's broadcast all confirm the remarks 13. The UN transcript is machine-generated and is not an official record, so we quote sparingly.

Amodei told the Council that four years ago AI "could barely write a line of code"; that today it "writes most of the code at Anthropic"; and that within "one or two years, maybe less" it could reach what he calls "a country of geniuses in the data center" 1. He named two risks: misuse, for example by bioterrorists, and loss of control. He called it "the most important global security issue facing the world today" and said Anthropic "will slow down as much as necessary" 1. He made three proposals 1:

  • narrow agreements every member can support, such as a ban on using AI to make biological weapons;
  • evaluation and verification systems that keep pace, so states can see frontier capabilities and verify each other's commitments;
  • common global testing standards for loss-of-control and misuse risks, plus a notification system for AI incidents significant to global security.
From the video summary: Anthropic’s CEO told the UN AI is “the most important global security issue facing the world today”. Are you ready?
From the video summary

Context on the speaker

Amodei leads a company with a direct commercial interest in AI, and his claims about capabilities are his own. The "country of geniuses in a datacenter" phrase comes from his October 2024 essay "Machines of Loving Grace" 4. His timelines have sometimes run ahead of events. In March 2025 he said AI would be writing 90 percent of code within three to six months 5. Six months later, independent coverage found no sign of that across the software industry 6. He has since said the figure holds "within Anthropic and within a number of companies that we work with" 7.

His record on oversight is concrete and checkable. In a September 2026 essay he called on the industry to "pace the frontier" 8. In October Anthropic onboarded six METR staff as embedded evaluators, with access to internal tools roughly on par with its own risk teams. METR can publish its findings, and Anthropic can't veto publication but can require redactions in predefined categories 9. Anthropic's Responsible Scaling Policy, now version 3.4 (effective 8 July 2026), ties safeguards to "AI Safety Levels". The company says its higher levels remain "largely undefined" 10.

The Council has been here before. Its first formal meeting on AI was on 18 July 2023 11. The Republic of Korea's president chaired an open debate in September 2025 12. In August 2025 the General Assembly set up the 40-member Scientific Panel and an annual Global Dialogue on AI Governance. Both are limited to the non-military domain 13.

Download the PDF. Give it to your agent.

Have your agent study our white papers alongside you.

Free download

Want the full version?

Leave your email and we’ll send it.

Adapted from “Things to Know: What the July 2026 Hugging Face Incident Teaches About AI Checking AI”, a Monarchy briefing (October 2026). Sources are listed below; where a source is quoted, the quotation is reproduced as published.

Sources

  1. United Nations, Security Council 10228th meeting, "Artificial intelligence and international security," 23 September 2026, transcript (automatic speech recognition; not an official record). https://transcripts.un.org/en/sc/10228
  2. UN Meetings Coverage SC/16462, "Ongoing Efforts to Create Powerful AI a 'Race Where Everyone Loses', UN Expert Tells Security Council…," 23 September 2026. https://press.un.org/en/2026/sc16462.doc.htm ; UN Audiovisual Library: https://media.un.org/avlibrary/en/asset/d363/d3637151
  3. UN Web TV, "Dario Amodei (CEO of Anthropic) on Artificial intelligence and international security," 23 September 2026. https://webtv.un.org/en/asset/k1v/k1vmsgetgo ; C-SPAN, "Artificial Intelligence CEOs Brief the U.N. Security Council on the Risks of AI." https://www.c-span.org/program/united-nations/artificial-intelligence-ceos-brief-the-un-security-council-on-the-risks-of-ai/685634
  4. Dario Amodei, "Machines of Loving Grace," October 2024. https://www.darioamodei.com/essay/machines-of-loving-grace
  5. Council on Foreign Relations, "CEO Speaker Series With Dario Amodei of Anthropic," 10 March 2025 (transcript). https://www.cfr.org/event/ceo-speaker-series-dario-amodei-anthropic
  6. Futurism, "Exactly Six Months Ago, the CEO of Anthropic Said That in Six Months AI Would Be Writing 90 Percent of Code," 10 September 2025. https://futurism.com/six-months-anthropic-coding
  7. Business Insider, "Anthropic CEO says 90% of code written by teams at the company is done by AI…" (Dreamforce conversation with Marc Benioff), October 2025. https://www.businessinsider.com/most-anthropic-teams-coding-with-claude-ai-not-replacing-humans-2025-10
  8. Dario Amodei, "We Must Pace the Frontier," September 2026. https://darioamodei.com/post/we-must-pace-the-frontier
  9. Anthropic, "An update on independent evaluators," 15 October 2026 (updated 16 October). https://www.anthropic.com/news/an-update-on-independent-evaluators
  10. Anthropic, "Responsible Scaling Policy" (current version 3.4, effective 8 July 2026). https://www.anthropic.com/responsible-scaling-policy ; "Responsible Scaling Policy Version 3.0," 24 February 2026. https://www.anthropic.com/news/responsible-scaling-policy-v3
  11. UN Meetings Coverage SC/15359, Security Council debate on artificial intelligence, 18 July 2023. https://press.un.org/en/2023/sc15359.doc.htm
  12. UN Audiovisual Library, Security Council 10005th meeting, "Artificial intelligence and international peace and security," 24 September 2025. https://media.un.org/avlibrary/en/asset/d346/d3460317
  13. UN General Assembly, resolution A/RES/79/325, adopted 26 August 2025. https://www.un.org/pga/wp-content/uploads/sites/110/2025/12/A_RES_79_325-EN.pdf

More from Monarchy

All articles →