The UK and South Korean governments, along with leading technology companies like Amazon and Google, have pledged to establish frameworks for assessing and mitigating risks associated with advanced AI models. This initiative builds on the Bletchley Declaration and aims to enhance global understanding of AI safety.
On May 21, 2024, the UK and South Korean governments announced new voluntary commitments on AI safety from leading technology companies. This announcement took place ahead of a two-day global AI summit held in Seoul. The companies involved include Amazon, Google, Meta, Microsoft, OpenAI, xAI, and Zhipu AI, among others. These companies have pledged to establish frameworks for assessing and mitigating risks associated with their advanced AI models.
This initiative builds on the Bletchley Declaration made at the inaugural AI Safety Summit hosted by UK Prime Minister Rishi Sunak in November 2023. Under this agreement, companies committed to halting the development of AI models if severe risks cannot be mitigated. They also agreed to outline thresholds for tolerable risks and implement transparent measures for mitigating these risks.
Key figures such as UK Science Secretary Michelle Donelan emphasized the importance of these voluntary commitments and noted the progress achieved since the Bletchley Summit. The commitments include publishing how companies will measure risks and the actions they will take to address those risks.
Additionally, ten nations and the European Union signed the “Seoul Statement of Intent toward International Cooperation on AI Safety Science.” This agreement aims to establish an international network of AI Safety Institutes to share research and enhance global understanding of AI safety.
Prime Minister Rishi Sunak co-chaired a virtual meeting with South Korean President Yoon Suk Yeol, attended by leaders and key figures from AI and tech firms, including Elon Musk and Eric Schmidt. The summit concluded with a joint commitment to continue collaboration on AI safety.
