Sovereign AI Programs

Corrected May 19, 2026Fresh19d

A country-by-country announced-versus-deployed reconciliation, joined against primary government procurement disclosures, grid-queue interconnection filings, and a licensed fund/LP database’s LP→fund chains. Headline commitments decompose into government-only disbursement versus private/FDI/credit-instrument padding, with the per-actor public-records traceability rate shown adjacent. Full supply-chain position by country at Countries; access constraints at Export Controls.

A country-by-country reckoning of announced commitments against deployed compute, with government-only spending separated from the headline figures that quietly conflate it with private pledges, foreign direct investment, and credit lines. Supply-chain position by country at Countries; access constraints at Export Controls.

Sovereign AI Programs

Corrected May 19, 2026Fresh19d

The brief

One screen, plain English: how far announced sovereign-AI money outruns deployed compute, what the government-only figures actually are, and why an em-dash is a data gap and never a zero — the finding without the dashboard.

The finding

Announced sovereign-AI money outruns what governments will spend, by a wide margin.

Across 34 national programs, the headline commitments add up to $1.1T. Strip out the private pledges, foreign direct investment, and credit lines that get folded into those announcements, and the money that actually sits in government budgets is $182.0B — about 83% of the headline is not government money at all. Of that government-only total, $5.7B has been confirmed spent — a 3.1% execution rate.

The typical program has disbursed 27.0% of what it announced (the median across programs that report both). Smaller commitments tend to clear at higher rates; the largest announcements drag the longest tails of unfunded headline.

Announced$1.1T

the headline figure, across 27 programs — the number that gets reported

Government-only$182.0B

what government budgets actually carry — the rest, about 83%, is private pledges, foreign investment, and credit lines wearing the government's banner

Disbursed$5.7B

money confirmed spent, across 13 programs that report it

What the numbers do and don’t say

An em-dash is a data gap, never a zero.Where a figure is not publicly available the record shows “—”; it is never treated as if the program committed nothing. Several programs (China, Indonesia, Poland, and others) publish no financial figures and are counted as unknown, not empty.
Government-only is the metric this rests on. It is decomposed for 27 of 34 programs where the source records permit; the remainder fall back to the headline, so the government-only total is a floor, not a ceiling.
Disbursed is not the same as deployed. Money confirmed spent is not yet compute in the ground; where a facility actually exists it shows up as a primary-sourced site, tracked separately.

The per-country reconciliation, execution tracker, and hardware-dependency reach are on the Programs view; the capital vehicles behind the money on Funding; the anatomy of the gap on Inflation.

Sovereign AICompute & Capability

The full announced-versus-deployed instrument behind this brief

Sovereign AI · Inflation

The anatomy of the announced-versus-deployed gap

Sovereign AI · Funding

The capital vehicles behind the money

Built for the AI governance research community