Artificial intelligence (AI) has moved closer to the epicentre of international security discussions after leaders of major AI companies called for stronger global cooperation and safeguards during a United Nations Security Council meeting held in New York.
Sam Altman, who is the chief executive of OpenAI, Anthropic CEO Dario Amodei and Hugging Face co-founder Clément Delangue, addressed the council on September 23 this year. They cautioned that if we do not manage the development and use of increasingly capable AI systems properly, they could pose serious risks. Amodei said poorly managed AI could at the end cause risks to humanity, while Altman encouraged governments to work with technology companies on common safety measures.
The meeting comes as governments struggle to establish common rules for technology that is developing faster than many existing regulatory systems. Altman has called for benchmarks that could measure AI capabilities and assess the safeguards companies introduce as their systems become more advanced.
However, international agreement continues to be difficult. The US administration’s representative at the Security Council, Michael Kratsios, explicitly rejected efforts by international bodies to establish centralized global governance of AI. Other governments have called for greater international cooperation and safeguards.
The discussions also follow the release of the first thematic brief by the UN’s Independent International Scientific Panel on AI on September 21. The brief examined a 2026 incident which involved AI agents and concluded that several conditions that are associated with potential loss of human control had occurred together in a real-world system. The panel said the incident raised questions about how AI agents are currently trained.
The UN has already started much larger work on international AI governance. Its General Assembly is preparing further dialogue on common approaches which covers aspects such as safety, accountability, human oversight, access to AI capabilities and cooperation between countries.
With AI development continuing at remarkable speed, the latest UN discussions highlight a central challenge for governments: creating safeguards that address genuine risks while allowing beneficial technologies to continue developing.
