Your next AI bet is
a geography problem.

Talent costs, infra maturity, and regulatory readiness vary wildly across borders. This map scores 193 countries on the factors that actually predict where AI investment pays off.

Stage 4: Leading
Stage 3: Prepared
Stage 2: Developing
Stage 1: Early stage
No data
193 countries
Sort by
1

United States

North America

87.7

+17.8

4
2

Japan

East Asia & Pacific

86.1

+11.8

4
3

United Kingdom

Europe & Central Asia

84.0

+17.3

4
4

Germany

Europe & Central Asia

84.0

+12.5

4
5

South Korea

East Asia & Pacific

83.6

+9.8

4
6

Canada

North America

82.9

+11.8

4
7

Australia

East Asia & Pacific

82.4

+9.1

4
8

Netherlands

Europe & Central Asia

81.6

+11.7

4
9

Denmark

Europe & Central Asia

81.2

+6.6

4
10

Sweden

Europe & Central Asia

80.9

-6.5

4
11

Spain

Europe & Central Asia

80.9

+15.3

4
12

Singapore

East Asia & Pacific

80.7

+6.2

4
13

France

Europe & Central Asia

80.6

-0.0

4
14

Italy

Europe & Central Asia

80.1

+8.9

4
15

Finland

Europe & Central Asia

79.8

+0.4

4
16

Norway

Europe & Central Asia

79.7

-8.9

4
17

Ireland

Europe & Central Asia

78.7

+9.0

4
18

Austria

Europe & Central Asia

77.8

-8.2

4
19

China

East Asia & Pacific

77.1

+6.4

4
20

Poland

Europe & Central Asia

76.5

+13.1

4

Source: Agence Française de Développement — AIIPI 2026 (CC-BY)

What the data says

Spending alone does not buy AI readiness. The countries that score highest invest in foundations — connectivity, talent pipelines, and institutions — not just compute.

6:1

The readiness gap is wider than the wealth gap

USA at 87.7, South Sudan at 14.2. If you are choosing where to deploy AI workloads, the infrastructure floor matters more than the GDP ceiling.

#19

Money cannot buy what China is missing

Billions in AI spending, yet ranked behind Poland. The index penalizes closed data ecosystems and weak inclusion — a warning for any top-down AI strategy.

9 / 15

Europe's quiet advantage

Nine of the top fifteen are European. Strong universities, high connectivity, and regulatory clarity beat raw VC volume. Your nearshore AI team might belong in Lisbon, not Palo Alto.

#28

India has the talent. Not the floor.

World-class engineers, but the index looks at the whole country — inclusion, connectivity, institutions. Great for outsourcing AI development. Harder for broad AI deployment.

~100

The "almost ready" trap

Half the world scores between 45 and 75 — prepared on paper, stalled in practice. The gap between "can pilot AI" and "can scale AI" is where most investment dies.

#52

Africa's leapfrog candidates

South Africa, Morocco, and Tunisia lead the continent. Mobile-first infrastructure and young populations create an AI adoption path that skips legacy IT entirely.

Source: AI Investment Potential Index (AIIPI) 2026, Agence Française de Développement. Published on data.gouv.fr under CC-BY.

What actually predicts AI success

The AIIPI trains machine-learning models against real AI investment flows — VC rounds, private equity, and M&A — then ranks which indicators best predict where capital lands. The results challenge common assumptions.

#1

Research output

The number of AI research articles published by a country's institutions is the single strongest predictor of investment. Not patents, not VC dollars — published research.

#2

Government effectiveness

Institutional quality and policy implementation capacity matter more than the policies themselves. A well-run government predicts AI investment better than having an AI strategy on paper.

#3

Data privacy & protection

Countries with strong data protection frameworks attract more AI investment, not less. Regulatory clarity builds the trust that enterprise AI deployments require.

#4

Mobile connectivity

The GSMA Connectivity Index — measuring digital accessibility for citizens — outweighs raw broadband speed. AI scales where the population can actually access it.

#5

Population

Market size still matters. Large domestic populations represent scalability and addressable market — the scope for AI technologies to generate returns.

Variable importance derived from Random Forest model trained on AI investment counts (VC, PE, and M&A), validated via 10-fold cross-validation. Source: AIIPI 2025 methodology paper.

Regional benchmarks

Average AIIPI scores by World Bank region. The gap between North America and Sub-Saharan Africa is not just a wealth story — it tracks infrastructure, governance, and data maturity.

North America
86.0
Stage 4
Europe & Central Asia
67.4
Stage 3
Middle East & North Africa
54.5
Stage 3
East Asia & Pacific
52.8
Stage 3
South Asia
49.7
Stage 2
Latin America & Caribbean
47.4
Stage 2
Sub-Saharan Africa
37.2
Stage 2

Regional averages from AIIPI 2025. Individual country scores available in the interactive map above.

How the index works

The AIIPI is published annually by the Agence Française de Développement (AFD), France's public development bank. It scores 193 countries across 19 indicators and 6 dimensions, using machine-learning models trained on real AI investment flows to determine what actually predicts where capital lands.

01

Economic environment

Market size, economic prosperity, energy access, and the sophistication of a country's productive structure.

Population GDP per Capita (PPP) Access to Electricity Economic Complexity Index Complexity Outlook Index

02

Governance

Democratic freedoms, institutional quality, policy implementation capacity, and investment climate stability.

Voice & Accountability Government Effectiveness Political Stability

03

Digital & physical infrastructure

Mobile connectivity, broadband maturity, and telecommunications networks that enable AI deployment at scale.

GSMA Connectivity Index Telecom Infrastructure Index

04

Human capital

Workforce skills, education levels, and the country's capacity for knowledge generation and AI research.

Human Capital Index AI Research Articles

05

Data governance

Government commitment to AI strategy and the strength of data privacy and protection frameworks.

National AI Strategy Data Privacy Score

06

Statistical performance

The quality of a nation's statistical system — how well it produces, manages, and shares data.

Data Use Data Services Data Products Data Sources Data Infrastructure

ML-derived weighting

Weights are not assigned by committee. The AIIPI trains Random Forest, XGBoost, Elastic Net, and Linear Regression models against actual AI investment flows, then derives feature importance from the best performer (Random Forest, validated via 10-fold cross-validation). Indicators that better predict real investment carry more weight.

Four investment stages

Stage 4 — Leading ≥ 76

Advanced AI ecosystems with exceptional investment potential.

Stage 3 — Prepared 51 – 75

Solid foundations, but gaps in one or more dimensions.

Stage 2 — Developing 26 – 50

Emerging capabilities. Targeted investment can unlock potential.

Stage 1 — Early < 26

Foundational gaps in infrastructure, governance, or human capital.