Global Markets
Home›Global Markets›Trade & Tariffs›OpenAI, Anthropic, and Hugging Face press UN for share…
OpenAI, Anthropic, and Hugging Face press UN for shared AI safety standards
OpenAI CEO Sam Altman urged national and international standards to measure AI capabilities and risks, while a Trump adviser argued against creating new global governance.
OpenAI, Anthropic, and Hugging Face executives have told the UN that the speed of AI development and the societal risks it poses require international coordination, including shared standards for evaluating and managing threats. BBC Business reports that Sam Altman called for common risk evaluation standards, echoed by Dario Amodei of Anthropic and Clement Delangue of Hugging Face.
Altman said AI decisions should not be made by labs in one location alone, and he urged “national and international” standards covering how to measure AI capabilities, assess risks, set safeguards, and maintain human oversight. He also argued that common standards would help countries compare evidence, verify compliance, and use a shared language for what is happening.
At the same UN conference, Michael Kratsios, a key technology advisor to U.S. President Donald Trump, rejected the idea that AI risk should lead to a pause in development or to new global governance structures. BBC Business notes that Kratsios said international dialogue should not drift toward global governance.
Amodei previously argued for slowing down as much as necessary to ensure each new AI release is safe, while Hugging Face has more recently been in the spotlight after admitting it was hacked by AI agents, or semi-autonomous AI bots, tied to OpenAI. BBC Business describes the episode as a “wake-up” moment.