OpenAI Urges US-Led Global Standards for AI Safety at United Nations

Concerns about the rapid pace of artificial intelligence development have been building for months, especially as new AI systems become more autonomous and capable. Researchers and policymakers have warned that without global coordination, advanced AI could pose serious risks to security and human control.

In September 2026, these concerns came to a head as top executives from leading AI companies, including OpenAI and Anthropic, prepared to address the United Nations. Their message was clear: the world needs shared, enforceable safety standards for AI, and the United States should play a central role in shaping them.

On September 21, a coalition of 22 countries released a joint statement stressing the need to keep AI "under human control" and calling for binding safety regulations. This group included many European nations, but the statement notably lacked the support of the United States and China, the two most influential countries in AI development.

Around the same time, Sam Altman publicly argued that outsiders, not just AI labs themselves, must help shape the rules for advanced AI. He emphasized that only shared, measurable standards could address the risks that cross national borders.

On September 23, Altman and Dario Amodei of Anthropic appeared before the UN Security Council. They warned that without international cooperation, the unchecked growth of AI could threaten humanity. Amodei even pledged that his company would slow down development if needed for safety.

At the same UN session, US officials rejected the idea of a single global AI regulator. Instead, they argued that each country should set its own rules, a stance that aligned with China's position. Both countries agreed only on limited cooperation, resisting calls for a broader slowdown or unified global governance.

As of late September 2026, there is strong momentum among many world leaders and industry executives for international AI safety standards. However, the two most influential AI nations, the US and China, have not agreed to binding global rules and prefer to regulate domestically.

OpenAI continues to push for a US-led approach to global standards, warning that fragmented national rules could make it harder to address cross-border risks. The debate now centers on whether meaningful international oversight can be achieved without the buy-in of the US and China.

World leaders and AI executives are expected to keep pressing for international cooperation in upcoming UN sessions. The key question is whether the US or China will shift their positions, or if global safety standards will move forward without their full participation.