Middle Powers in an Age of AI Superpower Dominance

By Anna Hope and Hassan Damluji

AI is concentrating power fast. A small number of states and firms could soon act with effective impunity, setting the terms on which the rest of the world’s economies, public services and security depend. The only way to prevent this is to address the concentration of power directly, and Middle Powers are key to this effort. 

Middle powers combine a strong need for effective international cooperation with the collective ability to act as rule-makers. This consultation paper argues they are uniquely placed to prevent the emergence of what it calls “AI imperialism” a structural outcome in which a handful of states and firms control the inputs on which everyone else’s economy, public service infrastructure and defence depend.

The paper sets out five tests for non-dependence in the AI age: data ownership, operational independence, access to competitive technology, capacity, and hardware access. No country other than the US or China can meet these tests alone. But pooled together, a coalition of Middle Powers could achieve collectively what none can achieve individually. They must decide between cooperation versus dependency.

This paper is intended to build consensus on the scale of the risk and a response that can mitigate it.