1 Introduction 3

2 The AI duopoly and the N‑2 problem 5

3 AI sovereignty: two literatures and a synthesis 7

4 Measuring hard and programmable sovereignty 11

5 Coordinated specialisation: from individual insufficiency to coalition sufficiency 16

6 Institutional precedents and their limits 19

7 Policy implications 21

8 Conclusion 23

References 24

Appendix: Coalition feasibility, supporting data 27