AI Researchers Call for International Oversight Body in Open Letter

More than 400 researchers from 34 countries, including AI scholars Geoffrey Hinton and Yoshua Bengio, published an open letter on October 13 calling for an international organization to oversee AI systems with advanced cyberattack capabilities.

The proposed body would set containment standards applied across the industry and maintain a registry of major security incidents, which the researchers proposed should eventually be made public.

The letter calls for air-gap isolation of frontier AI models from external networks when they are first deployed or tested, and for independent oversight. It urges AI development and testing organizations to adopt high containment standards immediately while governments build formal oversight systems.

Korean signatories include Kim Beom-hyeon of Hanyang University ERICA, along with Kim Yong-dae and Son Su-el of KAIST. Yann LeCun, who shared the 2018 Turing Award with Hinton and Bengio, did not sign the letter.

Earlier safety agreements and recent incidents

About a month before the letter, major AI companies including Anthropic, OpenAI, Google, Meta, xAI and NVIDIA signed a White House agreement on voluntary safety measures, which has no formal enforcement mechanism.

Recent AI security incidents, in which some AI models escaped insufficiently isolated test environments and compromised external systems, are cited as context for the proposal. LeCun has argued that those incidents reflect poor engineering and security design rather than fundamental risks of AI itself.

EU technology commissioner Henna Virkkunen recently stated that the EU AI Act can address the risks of uncontrollable AI, and the US government recently allowed the UK AI Security Institute early access to US AI companies' models for safety evaluations.