Supporting case
A Global South comparison shifts attention from country alignment to the capabilities governments actually obtain
What happened
Sabeen Dhanani’s October 21 CSIS analysis compares AI partnerships beyond a simple choice between the United States and China. It examines combinations of computing access, public digital infrastructure, locally useful models, skills and governance support, with India and Gulf states among the alternative partners. It is a comparative policy analysis rather than a new causal study or a newly executed international agreement.
Its useful question for The Race is what a recipient government retains: access rented from a vendor, ownership and operating expertise, reusable public infrastructure, or a continuing dependence on another country’s chips and cloud. Having a facility in another country does not answer those questions. The analysis provides a framework for comparing arrangements, not proof that every described partnership disperses power.
Compare what governments retain from partnerships: ownership, operating knowledge, reusable public infrastructure and choice, alongside hardware and cloud dependencies. Geographic diversification alone does not establish dispersion.
Dates and implementation stage
Development/status: 2026-10-21; source publication 2026-10-21
