OpenAI and Anthropic CEOs Warn UN Security Council of AI Risks

AI industry leaders call for global oversight

The chief executives of two leading artificial intelligence companies appeared before the United Nations Security Council on September 23, 2026, delivering stark warnings about the technology they are building and urging international action to establish controls.

Dario Amodei, CEO of Anthropic, told the Security Council that "if managed poorly, I even believe AI could be a risk to humanity as a whole." Sam Altman, who leads OpenAI, echoed the concern: "We could lose control of the future to AI."

Both executives argued that the UN Security Council should establish safeguards to prevent AI systems from becoming too powerful to control. They also emphasized the need to prevent the concentration of AI capabilities in a single company or nation.

The testimony came one day after UN Secretary-General Antonio Guterres raised concerns about "killer robots." According to the Associated Press, which first reported the details, experts and European leaders indicated such autonomous weapons systems are either already in use or dangerously close to deployment.

Why it matters

The appearance of major AI company leaders before the Security Council represents a significant moment in the global governance debate around artificial intelligence. Unlike typical industry lobbying for lighter regulation, these executives are actively requesting oversight—a recognition that the technology's trajectory may outpace any single company's or country's ability to manage risks. The Security Council setting also frames AI as a matter of international security rather than purely economic or technological policy, potentially opening pathways for binding international agreements similar to nuclear non-proliferation frameworks.

Central question: human or machine control

The Security Council session focused on a fundamental question confronting world leaders: who should make life-and-death decisions as AI systems become more capable—humans or machines?

The meeting brought together technology leaders and policymakers to examine whether AI's expanding capabilities and reach require formal controls, who should implement them, how such governance could function, and whether intervention might already be arriving too late.

The Associated Press reported these details from the UN Security Council meeting at the 81st session of the United Nations General Assembly.