OpenAI urges US to lead global AI standards effort

OpenAI called Monday for the United States to lead a global effort to set standards for developing artificial intelligence, as chief executive Sam Altman prepares to brief the United Nations Security Council this week.

The San Francisco-based AI company said it supported international cooperation on safety and security standards that could reduce risks from AI development. In a blog post, OpenAI said the United States was well placed to lead because its AI industry is at the technical frontier.

OpenAI also said recursive self-improvement, a technique in which software teaches itself and improves autonomously, should not be pursued unless it can be done safely. Critics fear such systems could have catastrophic outcomes if humans cannot intervene.

The company linked the concern to a July security breach involving Hugging Face, a site used by AI developers to store and share code. OpenAI said the incident involved unreleased models undergoing testing and was not a direct result of recursive self-improvement, but showed risks that could become more severe without safeguards and alignment.

Altman is expected to brief the UN Security Council on Wednesday. The meeting is being organized by France, which holds the rotating presidency of the 15-member council, during the same week as the UN General Assembly.

UN Secretary-General Antonio Guterres called Wednesday for coordinated international action on AI risks, saying national action was essential but global coordination was also indispensable.

The debate over AI safety has grown after security incidents involving models from OpenAI, Anthropic and Google, and after industry figures called for caution or a slowdown in AI development. President Donald Trump has dismissed warnings about AI risks as a “hoax.”