Beyond the “Bletchley declaration” and leaders’ speeches at the AI summit (Report, 2 November), the lack of a global AI governance arrangement is starkly absent. It would seem that we now have three (rival) attempts to coordinate and lead – the US Artificial Intelligence Safety Institute based in Washington DC, the Global Partnership on Artificial Intelligence based in Paris and the new London-based AI Safety Institute. Each will try to establish focality and authority.
All will undoubtedly seek to attract the best minds and create legitimacy through quality of output. They will fail to establish this legitimacy unless they make the difficult choice of being set up as an intergovernmental organisation. Only by doing this will the world’s major powers, including China, be represented in a managed and fair way.
