Leaders of two major artificial intelligence companies have warned the United Nations that advanced AI could create serious risks if governments and developers
fail to control it. OpenAI chief executive Sam Altman and Anthropic chief executive Dario Amodei spoke at a UN Security Council meeting held during the 81st UN General Assembly.
Amodei said AI could become a risk to humanity as a whole if it is managed poorly. Altman warned that people could lose control of the future to AI. Their comments were warnings about possible outcomes, not evidence that such an event is certain.
Calls for testing and safeguards
Both executives urged governments to develop safeguards before more capable systems arrive. Amodei called for common testing standards and ways to verify what powerful models can do. He also highlighted the danger that AI could help people develop biological weapons.
The executives also warned against allowing control of advanced AI to become concentrated in one company or country. The discussion included representatives from governments and technology companies, including Hugging Face chief executive Clément Delangue.
Countries remain divided
British Foreign Secretary Ed Miliband said the United Kingdom would put AI controls at the centre of G20 discussions next year and work toward global standards. US science adviser Michael Kratsios rejected calls to pause development or establish a new global governance structure.
Chinese UN Ambassador Fu Cong supported countries choosing their own development paths and backed open-source models. He also warned against digital dominance by a small number of actors.
France convened the meeting. French Foreign Minister Jean-Noël Barrot compared the challenge of governing AI with the international response to atomic technology. However, the meeting did not produce a binding agreement. The warnings therefore remain part of an ongoing policy debate over how to balance safety, competition and access.
