Executive Summary

Artificial intelligence presents the greatest transformation in human affairs since the industrial revolution, and it may be greater still. It promises material and informational abundance: machines to carry out the dangerous and arduous, leaving humanity wealthier, safer, and freer as a result. Yet challenges come with that: AI looks likely to upend the economic and industrial strategies of all countries. It may erode or reinforce the informational advantage that has underwritten the state's claim to govern. And it may throw off or bolster the favourable balance between defense and offense that secures a state’s ability to protect its citizens from harm.

The nations with the most to lose and gain from this transformation will be the middle powers: nations with significant institutional and industrial capacity but without frontier AI development capability. Unlike great powers, they do not control whether and how AI gets built, and whether it rips through society or not. Yet unlike the least developed nations, middle powers possess institutional substrate—functioning economies and bureaucracies. They have something to transform. Unburdened by the need to win a race they were never going to win, they are free to—but must—ask different questions on how to deploy the assets they still have to secure the future they want.

The race that remains open, the race where the middle powers can compete and even lead, is the race to favourably adapt frontier artificial intelligence into their governments and economies. This fact cannot stand as an excuse to remain inactive, but instead has to be understood as a generational effort. It will require finding answers to encompassing and complicated questions that have no comfortable answers:

• Access architecture: What contractual, technical, and institutional arrangements can durably guarantee middle powers' access to imported frontier AI?
• Institutional transformation in practice: How can the public sector be reshaped to be amenable to AI-driven transformation? What separates the risks of falling behind and the opportunities for leapfrogging?
• Economic bottlenecks: Where along the flow of AI through the global economy can a given middle power occupy a bottleneck valuable enough to safeguard its fiscal and economic prospects?
• Redistribution mechanics: How can middle powers support their citizens when AI-driven revenue shifts simultaneously from labour income to capital returns and from domestic economies to the countries where AI is built?

Middle powers that fail will do so in familiar ways with unfamiliar dramatic consequences: allowing the legitimacy of the state to erode, taking half measures and easy roads, shutting their eyes and closing their borders to a global phenomenon, and waiting for a sovereign solution that never arrives. Those that move with clarity and resolve will shape the transformation.