Sam Altman Addresses UN Security Council on AI Risks and Standards

On September 23, 2026, Sam Altman, CEO of OpenAI, addressed the United Nations Security Council, stating that a 10% risk of catastrophe from AI is unacceptable. He emphasized that even a 1% risk is too high, underscoring the urgency for establishing complementary national and international frontier AI standards. Altman noted significant advancements in AI capabilities over the past three summers: models were proficient in grade-school math three summers ago, excelled in high-school math competitions two summers ago, and achieved gold-medal status in international math competitions last summer. In recent weeks, he observed rapid progress in AI models, reinforcing his claim that 'none of these levels are remotely acceptable.' He articulated a vision for AI's future, suggesting it could lead to either a new Renaissance of creativity or an Industrial Revolution characterized by upheaval. Altman concluded that we are at a critical juncture, necessitating immediate action to harness AI's potential responsibly.