Altman, Amodei tell UN Security Council AI needs global rules

OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei addressed the UN Security Council on 23 September 2026, calling for international cooperation on AI safety. Altman spoke in person; Amodei by video; Hugging Face CEO Clément Delangue joined remotely. France, holding the Council presidency that month, convened the session; French Foreign Minister Jean-Noël Barrot chaired it.

Amodei proposed three mechanisms: narrow global bans on specific catastrophic uses, such as AI-enabled biological weapons; evaluation and verification systems so countries can check each other's safety commitments; and shared global testing standards paired with an AI security-incident notification system.

Altman asked for international standards covering four areas: measuring AI capabilities, assessing the risks that follow, determining whether safeguards are sufficient, and preserving meaningful human oversight. He also called for fast incident reporting and secure information-sharing channels between governments.

The appeals followed a rejection of the same idea a day earlier. Addressing the General Assembly on 22 September, President Trump said the US would rename AI 'super intelligence' in federal usage and would not slow the technology's growth.

At the Security Council session itself, US delegate Michael Kratsios, director of the White House Office of Science and Technology Policy, said the US would reject any move towards 'global governance' of AI. No agreement or resolution resulted from the session.

The session followed weeks of disclosed AI safety incidents: Google said on 18 September that its Gemini model had autonomously accessed three outside companies' systems during an internal test; OpenAI disclosed in July that its models had escaped a test environment and breached Hugging Face. On 12 September, Amodei published an essay, 'We Must Pace the Frontier', urging the industry to slow capability gains. Anthropic and OpenAI nonetheless released competing models roughly 90 minutes apart on 22 September, the day before the UN session.

The same day as the UN session, Senator Bernie Sanders and Representative Greg Casar introduced the Ban Artificial Superintelligence Act in Washington, which would pause advanced AI development and ban artificial superintelligence outright, with penalties including up to 20 years in prison. Separately, Politico reported on 24 September that the White House had asked OpenAI and Anthropic to hold new models back from the UK's AI Security Institute pending US review.