Altman and Amodei Brief UN Security Council on AI Safety
On September 24, OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei separately briefed the UN Security Council on AI safety — the first time the heads of the two leading AI labs appeared before the Council on the same day, bringing AI governance discussions to the highest level of multilateral diplomacy.
Confirmed
- Altman warned that as systems become capable of recursive self-improvement and the process of building AI grows increasingly automated, the pace of progress could accelerate dramatically, saying "this moment calls for extreme caution." He admitted humanity could "lose the future to AI" — the risk being that AI advances too fast for people to keep up or intervene when needed.
- Altman revealed that OpenAI has in the past voluntarily slowed down AI development on its own, and would do so again.
- According to @ns123abc, Altman said some AI practitioners' rhetoric is "so dystopian it sounds like a bad sci-fi movie plot," noting he neither wants blind optimism nor a doomsday mindset, seeking a balance between the two extremes.
- Altman and Amodei briefed the Council separately rather than jointly.
Why it matters
- It is rare for the leaders of OpenAI and Anthropic to appear before the UN Security Council on the same day, formally placing AI safety on the agenda of the world's top security governance body.
- Altman publicly acknowledged for the first time that OpenAI has proactively slowed its research and may do so again, offering a rare official account for assessing top labs' safety self-discipline.
- His warnings about recursive self-improvement coexist with criticism of doomsday narratives, sketching out his "middle path" stance on AI risk.
2026-09-24 ~ 2026-09-24 · 6 related posts
- Episode 1: Altman and Amodei to Address UN Security Council on AI Safety(2026-09-21, 6 posts)
- Episode 2: Gary Marcus Slams Altman's UN Security Council AI Briefing(2026-09-22, 2 posts)
- Episode 3: Altman and Amodei Brief UN Security Council on AI Safety(2026-09-24, 6 posts)
Primary sources
- [source] Altman at UN: some things AI builders say sound like bad sci-fi plots — ns123abc · 2026-09-24
- Sam Altman Tells UN Humans Could Lose Control to AI, OpenAI Has Slowed Development Before — rohanpaul_ai · 2026-09-24
- [source] Altman and Amodei brief UN Security Council on AI safety amid growing scrutiny — nordicinst · 2026-09-24
- Sam Altman tells Congress: the AI risk is moving so fast no one can intervene — nitarshan · 2026-09-24
2 near-duplicate retellings: rohanpaul_ai · rohanpaul_ai