OpenAI CEO Urges UN Security Council to Establish Global AI Safety Mechanism

Sam Altman, CEO of OpenAI, delivered a roughly 10-minute speech at the UN Security Council on September 24, 2026, calling on countries to set up AI safety safeguards through the council to prevent AI power from being monopolized by any single company or nation. He stressed that AI systems must remain under human control and should benefit humanity by empowering individuals.

Altman praised AI's potential to improve productivity while warning about the risks of misuse. He advocated a "middle path" in the safety debate based on three principles: AI systems must always remain under human control, AI must benefit humanity—from assisting doctors in treating patients to helping entrepreneurs—and the technology must empower individuals. He acknowledged concerns about existential risks raised by some researchers but cautioned against falling into "doomsday" narratives.

At the same meeting, Dario Amodei, CEO of Anthropic, joined via video link and reiterated his calls for stronger global cooperation and regulation. He proposed international agreements to ban the use of AI in developing biological weapons and to create global standards for testing new AI models. Clement Delangue, CEO of Hugging Face, also spoke, urging greater transparency from private companies developing cutting-edge AI models and arguing against fear-based messaging.

The remarks came amid tightening product safety reviews for AI companies. Earlier in September 2026, Anthropic researchers had warned that the technology could pose risks to humanity within a decade, fueling ongoing public concern. Altman and Amodei, along with other AI founders, had previously called for a slower pace of AI development and stronger international cooperation on AI safety standards.