AI adoption varies dramatically across countries. This page shows real per-capita AI (Claude.ai) usage from the Anthropic Economic Index (May 2026) — a usage-based measure grounded in observed behaviour, not forecasts.
Data as of Jun 2026
Countries tracked
201
With measurable usage
121
Top usage index (Australia)
6.40
Global AI Adoption — World Map
Two lenses are available via the layer toggle: Claude.ai usage (per-capita observed interactions, Anthropic Economic Index Apr–May 2026 — availability-biased; China and restricted markets appear grey) and GenAI diffusion (Microsoft AIEI Q2 2026, % of working-age population using GenAI across 147 countries, China included). The two metrics use different denominators and cannot be merged — see Data & Sources.
AI job demand uses Indeed Hiring Lab job-posting share data for 9 economies, latest month.
GenAI diffusion leaders · Microsoft AIEI Q2 2026
United Arab Emirates73.3%
Singapore64.3%
Ireland49.9%
% working-age population using generative AI (147-country survey).
🇨🇳
China — Proxy Context
Claude layer: grey
Claude.ai is unavailable in mainland China, so it appears grey on the Claude.ai usage layer and is excluded from the per-capita usage index. On the GenAI diffusion layer, China does appear — Microsoft AIEI estimates ~17.5% of working-age adults used GenAI in Q2 2026. Note that Western telemetry likely undercounts domestic apps (Doubao, Kimi, etc.) — CNNIC’s survey implies ~43% penetration. The native-ecosystem figures below use different measurement approaches and denominators and are not merged into either index.
CNNIC · Jun 2025
515M
Generative-AI users
QuestMobile · H1 2025
680M
Mobile-AI MAU
Doubao (QuestMobile) · Dec 2025
226M
App MAU
Microsoft AIEI · Q2 2026
~17.5%
GenAI diffusion (working-age pop.)
These proxies use different measurement methods (government survey, app-market scan, product MAU) and cannot be summed or directly compared to each other. The usageIndex (Claude.ai interactions per working-age capita) and diffusionPct (Microsoft AIEI survey %) use entirely different denominators — do not merge them. See the Data & Sources page for full provenance details.
Supplemental proxy evidence
AI Adoption Signals
Heterogeneous proxy evidence from surveys, app-market measures, open-model activity, developer sources, and research activity. These signals are not merged into the Claude usage index.
Caveat: These metrics use different denominators and collection methods. Do not merge them into the Anthropic usageIndex without explicit normalization and labeling.
Collected families
11
Visualized families
11
Future candidates
8
enterpriseAdoptionMetrics
Enterprise AI survey shares
Business survey measures reporting respondent share using at least one AI technology. Comparable only within each source family.
Denmark (OECD) · Denmark42.0%
Finland (OECD) · Finland37.8%
Sweden (OECD) · Sweden35.0%
Belgium (OECD) · Belgium34.5%
Luxembourg (OECD) · Luxembourg33.6%
Netherlands (OECD) · Netherlands33.2%
Austria (OECD) · Austria29.9%
Norway (OECD) · Norway28.9%
Euro area (EA11-1999, EA12-2001, EA13-2007, EA15-2008, EA16-2009, EA17-2011, EA18-2014, EA19-2015, EA20-2023, EA21-2026) (Eurostat): 14.4%European Union - 27 countries (from 2020) (Eurostat): 13.5%OECD (OECD): 20.3%
Source: OECD
Period: 2025
Caveat: Enterprise adoption survey metric; comparable across listed European reporting countries, but not comparable to consumer MAU or Anthropic Claude.ai usageIndex.
Survey-based respondent share for generative AI use by individuals, shown as proxy evidence rather than product telemetry.
Norway · Norway56.3%
Denmark · Denmark48.4%
Switzerland · Switzerland47.0%
Estonia · Estonia46.6%
Finland · Finland46.3%
Ireland · Ireland44.9%
Netherlands · Netherlands44.7%
Greece · Greece44.1%
OECD: 36.8%
Source: OECD
Period: 2025
Caveat: Individual generative-AI usage survey metric; comparable across listed OECD/partner reporting countries, not comparable to enterprise adoption or product MAU.
U.S. Census business technology measures shown separately because geography, vintage, and denominator differ from global usage measures.
United States (2018 pre-GenAI ABS) · United States
3.0%
2018
District of Columbia (2018 pre-GenAI ABS) · District of Columbia
8.4%
2018
Delaware (2018 pre-GenAI ABS) · Delaware
6.9%
2018
Colorado (2018 pre-GenAI ABS) · Colorado
3.9%
2018
Hawaii (2018 pre-GenAI ABS) · Hawaii
3.9%
2018
Tennessee (2018 pre-GenAI ABS) · Tennessee
3.7%
2018
Nevada (2018 pre-GenAI ABS) · Nevada
3.6%
2018
California (2018 pre-GenAI ABS) · California
3.5%
2018
Georgia (2018 pre-GenAI ABS) · Georgia
3.5%
2018
Source: U.S. Census Bureau
Period: 2018
Caveat: U.S. employer-firm technology-use survey metric from 2018; predates the generative-AI wave and is not comparable to 2025 generative-AI usage surveys. 2018 pre-GenAI ABS.
Stack Overflow overall AI-tool response distributions, with each survey question kept separate.
ai_threat: No68.1%
ai_select: Yes61.8%
ai_sent: Favorable48.3%
ai_acc: Somewhat trust40.3%
ai_complex: Good, but not great at handling complex tasks32.7%
ai_complex: Bad at handling complex tasks31.3%
ai_acc: Neither trust nor distrust26.6%
ai_select: No, and I don't plan to24.4%
ai_sent: Very favorable23.6%
ai_acc: Somewhat distrust22.5%
ai_complex: Neither good or bad at handling complex tasks20.8%
ai_threat: I'm not sure19.9%
ai_sent: Indifferent18.7%
ai_select: No, but I plan to soon13.8%
ai_threat: Yes12.1%
ai_complex: Very poor at handling complex tasks11.9%
ai_acc: Highly distrust7.9%
ai_sent: Unfavorable5.2%
ai_complex: Very well at handling complex tasks3.3%
ai_sent: Unsure3.0%
ai_acc: Highly trust2.7%
ai_sent: Very unfavorable1.2%
Source: Stack Overflow / R4DS TidyTuesday
Period: 2024-05
Caveat: Developer survey proxy; country values are respondent shares, not population-representative national adoption rates. Overall distributions keep each survey question separate.
Country rows show Stack Overflow respondent yes-shares for AI-tool use, not population adoption.
Kenya respondent share · Kenya79.9%
China respondent share · China79.3%
Pakistan respondent share · Pakistan78.5%
Egypt respondent share · Egypt77.5%
Nigeria respondent share · Nigeria75.8%
Nepal respondent share · Nepal74.8%
Viet Nam respondent share · Viet Nam74.5%
Colombia respondent share · Colombia74.1%
Source: Stack Overflow / R4DS TidyTuesday
Period: 2024-05
Caveat: Developer survey proxy; country values are respondent shares, not population-representative national adoption rates. Country rows are respondent shares, not population adoption.
Country-level provider identity proxy from the OpenRouter public model catalog snapshot as of 2026-10-07. Model catalog counts and endpoint entries are separate lenses.
Caveat: Public catalog and endpoint availability only; not user traffic, usage, revenue, or national adoption. It is also not physical server location, training location, or a definitive national AI activity measure.
33 mapped model-provider entries across countries.
Endpoint entries
1,390
971 mapped endpoint entries kept separate from model counts.
Unknown/unmapped providers
56
46 model rows and 419 endpoint entries remain unmapped.
Top countries by model catalog count
ModelsEndpoint entries
Model catalog counts are the primary bars. Endpoint entries are shown as secondary bars and labels, without combining the two measures.
United States
243
691
China
135
177
France
26
55
Canada
6
7
Japan
4
4
South Korea
3
12
United Arab Emirates
2
8
Netherlands
0
11
United States: 243 Models, 691 Endpoint entries.
China: 135 Models, 177 Endpoint entries.
France: 26 Models, 55 Endpoint entries.
Canada: 6 Models, 7 Endpoint entries.
Japan: 4 Models, 4 Endpoint entries.
South Korea: 3 Models, 12 Endpoint entries.
United Arab Emirates: 2 Models, 8 Endpoint entries.
Netherlands: 0 Models, 11 Endpoint entries.
Country-level OpenRouter catalog proxy table
Country
Region
Model providers
Models
Endpoint providers
Endpoints
Top families
United States
North America
16
243
25
691
Gpt (96), Gemini (34), Claude (33)
China
Asia
12
135
11
177
Qwen3 (47), Glm (20), Deepseek (18)
France
Europe
1
26
1
55
Mistral (17), Ministral (4), Codestral (2)
Canada
North America
1
6
1
7
Command (5), North (1)
Japan
Asia
1
4
1
4
Fugu (3), Sakana (1)
South Korea
Asia
1
3
2
12
Solar (3)
United Arab Emirates
Middle East
1
2
2
8
Mercury (2)
Netherlands
Europe
0
0
1
11
—
Israel
Middle East
0
0
1
6
—
Joined country comparison
Global AI ecosystem comparison map
Joins OpenRouter model catalog footprint, GenAI diffusion, readiness scores, and adoption-readiness gaps into one country table. Filters keep catalog proxies separate from adoption/readiness metrics.
Country-level AI ecosystem comparison across model catalog footprint, readiness, diffusion, and gap quadrant.
Country
Models / endpoints
Readiness
Diffusion
Quadrant
United States
North America · USA
243 / 691
77.1
33.0%
Balanced Leader
China
Asia · CHN
135 / 177
63.5
17.5%
Latent Capacity
France
Europe · FRA
26 / 55
69.8
49.6%
Balanced Leader
Canada
North America · CAN
6 / 7
71.3
38.5%
Balanced Leader
Japan
Asia · JPN
4 / 4
73.3
24.7%
Latent Capacity
South Korea
Asia · KOR
3 / 12
72.7
40.6%
Balanced Leader
United Arab Emirates
Middle East · ARE
2 / 8
62.8
73.3%
Adoption Outpacing Readiness
Singapore
Other · SGP
0 / 0
80.1
64.3%
Balanced Leader
Denmark
Other · DNK
0 / 0
77.9
32.4%
Latent Capacity
The Netherlands
Europe · NLD
0 / 11
76.6
43.5%
Balanced Leader
Finland
Other · FIN
0 / 0
75.8
30.6%
Latent Capacity
Switzerland
Other · CHE
0 / 0
75.7
39.1%
Balanced Leader
New Zealand
Other · NZL
0 / 0
75.4
44.1%
Balanced Leader
Germany
Other · DEU
0 / 0
75.3
32.3%
Balanced Leader
Sweden
Other · SWE
0 / 0
74.8
37.3%
Balanced Leader
United Kingdom
Other · GBR
0 / 0
73.1
43.7%
Balanced Leader
Proxy caveat: OpenRouter is a public catalog/provider-identity footprint, not traffic, usage, demand, physical deployment, or national adoption. Readiness and diffusion use different denominators and should not be averaged with catalog counts.
Alignment lens
Adoption–Readiness Gap
Compares each country’s generative-AI diffusion percentile with its AI readiness percentile to surface where observed use and capacity are not aligned.
Caveat: Descriptive alignment only; the gap compares percentile ranks across two sources and is not a causal claim.
62.2% coverage of 201 mapped countries with both inputs.
Largest positive gap
Lebanon
Lebanon · +38.7 pctile
Largest latent capacity
Armenia
Armenia · -51.6 pctile
Readiness score vs. GenAI diffusion
Albania: Readiness 52.7, Diffusion 19.2%, Gap -8.1 pctile
Algeria: Readiness 37.0, Diffusion 13.8%, Gap +9.7 pctile
Angola: Readiness 26.0, Diffusion 11.4%, Gap +25.4 pctile
Argentina: Readiness 47.4, Diffusion 23.4%, Gap +12.9 pctile
Armenia: Readiness 49.3, Diffusion 7.7%, Gap -51.6 pctile
Australia: Readiness 72.7, Diffusion 40.7%, Gap +0.8 pctile
Austria: Readiness 72.5, Diffusion 35.4%, Gap -2.4 pctile
Azerbaijan: Readiness 47.1, Diffusion 18.7%, Gap +2.4 pctile
Bangladesh: Readiness 38.4, Diffusion 8.0%, Gap -23.8 pctile
Belgium: Readiness 67.2, Diffusion 40.3%, Gap +5.6 pctile
Benin: Readiness 36.3, Diffusion 10.4%, Gap -2.4 pctile
Bolivia: Readiness 37.7, Diffusion 13.3%, Gap +5.2 pctile
Bosnia and Herzegovina: Readiness 42.8, Diffusion 23.2%, Gap +21.0 pctile
Botswana: Readiness 41.3, Diffusion 15.0%, Gap +7.3 pctile
Brazil: Readiness 50.1, Diffusion 20.2%, Gap -4.0 pctile
Bulgaria: Readiness 57.7, Diffusion 30.7%, Gap +6.5 pctile
Burkina Faso: Readiness 31.2, Diffusion 10.4%, Gap +9.7 pctile
Burundi: Readiness 29.5, Diffusion 8.0%, Gap -1.2 pctile
Cambodia: Readiness 37.0, Diffusion 6.1%, Gap -25.8 pctile
Cameroon: Readiness 34.1, Diffusion 9.1%, Gap -4.8 pctile
Canada: Readiness 71.3, Diffusion 38.5%, Gap 0.0 pctile
Chad: Readiness 23.4, Diffusion 9.1%, Gap +11.3 pctile
Chile: Readiness 58.6, Diffusion 24.0%, Gap -11.3 pctile
China: Readiness 63.5, Diffusion 17.5%, Gap -33.1 pctile
Colombia: Readiness 48.9, Diffusion 25.9%, Gap +16.9 pctile
Costa Rica: Readiness 54.0, Diffusion 29.7%, Gap +8.9 pctile
Croatia: Readiness 58.2, Diffusion 27.2%, Gap -0.8 pctile
Czechia: Readiness 64.6, Diffusion 31.2%, Gap -2.8 pctile
Denmark: Readiness 77.9, Diffusion 32.4%, Gap -15.3 pctile
Dominican Republic: Readiness 46.9, Diffusion 25.8%, Gap +21.8 pctile
Ecuador: Readiness 44.2, Diffusion 20.3%, Gap +11.3 pctile
Egypt: Readiness 39.4, Diffusion 15.5%, Gap +11.3 pctile
El Salvador: Readiness 39.0, Diffusion 20.1%, Gap +23.4 pctile
Finland: Readiness 75.8, Diffusion 30.6%, Gap -19.4 pctile
France: Readiness 69.8, Diffusion 49.6%, Gap +11.3 pctile
Gabon: Readiness 32.3, Diffusion 15.7%, Gap +32.3 pctile
Gambia: Readiness 36.0, Diffusion 11.5%, Gap +8.1 pctile
Georgia: Readiness 53.0, Diffusion 21.6%, Gap -3.2 pctile
Germany: Readiness 75.3, Diffusion 32.3%, Gap -11.7 pctile
Ghana: Readiness 42.5, Diffusion 10.4%, Gap -18.6 pctile
Greece: Readiness 58.2, Diffusion 21.5%, Gap -15.3 pctile
Guatemala: Readiness 39.0, Diffusion 17.3%, Gap +14.9 pctile
Guinea: Readiness 32.4, Diffusion 10.4%, Gap +7.3 pctile
Guinea-Bissau: Readiness 26.5, Diffusion 10.4%, Gap +16.9 pctile
Guyana: Readiness 42.4, Diffusion 11.3%, Gap -12.5 pctile
Haiti: Readiness 26.8, Diffusion 8.9%, Gap +5.7 pctile
Honduras: Readiness 34.2, Diffusion 14.5%, Gap +21.8 pctile
Hungary: Readiness 56.3, Diffusion 33.3%, Gap +15.3 pctile
India: Readiness 49.3, Diffusion 18.5%, Gap -4.8 pctile
Indonesia: Readiness 51.6, Diffusion 14.8%, Gap -19.4 pctile
Iraq: Readiness 27.0, Diffusion 13.0%, Gap +28.2 pctile
Ireland: Readiness 69.3, Diffusion 49.9%, Gap +12.9 pctile
Israel: Readiness 72.5, Diffusion 38.7%, Gap -0.8 pctile
Italy: Readiness 62.1, Diffusion 31.4%, Gap +3.6 pctile
Ivory Coast: Readiness 36.6, Diffusion 13.7%, Gap +12.9 pctile
Jamaica: Readiness 43.4, Diffusion 25.0%, Gap +24.2 pctile
Japan: Readiness 73.3, Diffusion 24.7%, Gap -27.4 pctile
Jordan: Readiness 48.3, Diffusion 31.2%, Gap +27.8 pctile
Kazakhstan: Readiness 55.2, Diffusion 16.9%, Gap -24.2 pctile
Kenya: Readiness 44.5, Diffusion 8.9%, Gap -35.5 pctile
Kuwait: Readiness 46.1, Diffusion 22.6%, Gap +14.1 pctile
Kyrgyzstan: Readiness 42.6, Diffusion 10.0%, Gap -24.2 pctile
Laos: Readiness 33.0, Diffusion 8.4%, Gap -6.5 pctile
Lebanon: Readiness 41.8, Diffusion 27.7%, Gap +38.7 pctile
Lesotho: Readiness 35.5, Diffusion 9.9%, Gap -6.5 pctile
Liberia: Readiness 37.0, Diffusion 10.4%, Gap -6.5 pctile
Lithuania: Readiness 66.5, Diffusion 25.1%, Gap -16.9 pctile
Madagascar: Readiness 30.5, Diffusion 11.4%, Gap +18.1 pctile
Malawi: Readiness 34.0, Diffusion 11.4%, Gap +11.7 pctile
Malaysia: Readiness 63.2, Diffusion 22.6%, Gap -19.0 pctile
Mauritania: Readiness 23.3, Diffusion 10.4%, Gap +20.2 pctile
Mexico: Readiness 53.2, Diffusion 21.4%, Gap -5.6 pctile
Moldova: Readiness 48.1, Diffusion 19.1%, Gap +2.4 pctile
Mongolia: Readiness 48.4, Diffusion 17.9%, Gap -4.0 pctile
Morocco: Readiness 42.9, Diffusion 12.0%, Gap -9.7 pctile
Mozambique: Readiness 25.7, Diffusion 11.4%, Gap +26.2 pctile
Namibia: Readiness 42.0, Diffusion 15.4%, Gap +5.6 pctile
Nepal: Readiness 35.1, Diffusion 14.6%, Gap +21.8 pctile
New Zealand: Readiness 75.4, Diffusion 44.1%, Gap 0.0 pctile
Niger: Readiness 32.6, Diffusion 10.4%, Gap +6.5 pctile
Nigeria: Readiness 33.6, Diffusion 10.4%, Gap +4.8 pctile
Norway: Readiness 70.6, Diffusion 49.4%, Gap +9.7 pctile
Oman: Readiness 53.3, Diffusion 27.6%, Gap +9.7 pctile
Pakistan: Readiness 36.9, Diffusion 11.9%, Gap +5.7 pctile
Panama: Readiness 50.1, Diffusion 24.2%, Gap +6.5 pctile
Papua New Guinea: Readiness 29.0, Diffusion 7.6%, Gap -4.0 pctile
Paraguay: Readiness 41.0, Diffusion 12.6%, Gap -1.6 pctile
Peru: Readiness 49.1, Diffusion 17.3%, Gap -7.7 pctile
Philippines: Readiness 49.8, Diffusion 20.7%, Gap -0.8 pctile
Poland: Readiness 59.7, Diffusion 32.3%, Gap +6.1 pctile
Portugal: Readiness 64.6, Diffusion 27.4%, Gap -8.9 pctile
Qatar: Readiness 53.5, Diffusion 43.5%, Gap +28.6 pctile
Republic of the Congo: Readiness 27.7, Diffusion 9.1%, Gap +6.5 pctile
Romania: Readiness 58.4, Diffusion 18.0%, Gap -25.0 pctile
Rwanda: Readiness 43.7, Diffusion 7.6%, Gap -41.1 pctile
Saudi Arabia: Readiness 57.7, Diffusion 31.4%, Gap +10.1 pctile
Senegal: Readiness 39.6, Diffusion 14.4%, Gap +5.7 pctile
Serbia: Readiness 53.7, Diffusion 25.3%, Gap +1.6 pctile
Sierra Leone: Readiness 29.8, Diffusion 10.4%, Gap +12.1 pctile
Singapore: Readiness 80.1, Diffusion 64.3%, Gap -0.8 pctile
Slovakia: Readiness 59.2, Diffusion 27.1%, Gap -4.8 pctile
Slovenia: Readiness 63.4, Diffusion 30.1%, Gap -3.2 pctile
South Africa: Readiness 49.7, Diffusion 23.9%, Gap +6.5 pctile
South Korea: Readiness 72.7, Diffusion 40.6%, Gap +0.8 pctile
Spain: Readiness 64.8, Diffusion 45.1%, Gap +12.9 pctile
Sri Lanka: Readiness 43.6, Diffusion 7.6%, Gap -40.3 pctile
Suriname: Readiness 41.8, Diffusion 11.3%, Gap -10.9 pctile
Sweden: Readiness 74.8, Diffusion 37.3%, Gap -6.5 pctile
Switzerland: Readiness 75.7, Diffusion 39.1%, Gap -6.5 pctile
Tajikistan: Readiness 36.6, Diffusion 6.3%, Gap -23.4 pctile
Tanzania: Readiness 35.2, Diffusion 8.0%, Gap -13.3 pctile
Thailand: Readiness 53.6, Diffusion 13.4%, Gap -29.8 pctile
The Netherlands: Readiness 76.6, Diffusion 43.5%, Gap -4.4 pctile
Togo: Readiness 31.6, Diffusion 10.4%, Gap +8.9 pctile
Tunisia: Readiness 46.5, Diffusion 13.9%, Gap -8.9 pctile
Turkey: Readiness 54.0, Diffusion 18.9%, Gap -16.1 pctile
Uganda: Readiness 35.4, Diffusion 8.0%, Gap -14.1 pctile
Ukraine: Readiness 51.2, Diffusion 9.5%, Gap -46.0 pctile
United Arab Emirates: Readiness 62.8, Diffusion 73.3%, Gap +21.8 pctile
United Kingdom: Readiness 73.1, Diffusion 43.7%, Gap +2.4 pctile
United States: Readiness 77.1, Diffusion 33.0%, Gap -13.7 pctile
Uruguay: Readiness 54.9, Diffusion 25.7%, Gap 0.0 pctile
Vietnam: Readiness 48.2, Diffusion 27.9%, Gap +24.2 pctile
Zambia: Readiness 37.1, Diffusion 13.3%, Gap +6.0 pctile
Zimbabwe: Readiness 30.5, Diffusion 8.9%, Gap +0.8 pctile
Adoption outpacing readiness
Lebanon+38.7 pctile
Diffusion 27.7%Readiness 41.8
Gabon+32.3 pctile
Diffusion 15.7%Readiness 32.3
Qatar+28.6 pctile
Diffusion 43.5%Readiness 53.5
Iraq+28.2 pctile
Diffusion 13.0%Readiness 27.0
Jordan+27.8 pctile
Diffusion 31.2%Readiness 48.3
Latent capacity
Armenia-51.6 pctile
Diffusion 7.7%Readiness 49.3
Ukraine-46.0 pctile
Diffusion 9.5%Readiness 51.2
Rwanda-41.1 pctile
Diffusion 7.6%Readiness 43.7
Sri Lanka-40.3 pctile
Diffusion 7.6%Readiness 43.6
Kenya-35.5 pctile
Diffusion 8.9%Readiness 44.5
Balanced leaders
Singapore-0.8 pctile
Diffusion 64.3%Readiness 80.1
The Netherlands-4.4 pctile
Diffusion 43.5%Readiness 76.6
New Zealand0.0 pctile
Diffusion 44.1%Readiness 75.4
United Kingdom+2.4 pctile
Diffusion 43.7%Readiness 73.1
Switzerland-6.5 pctile
Diffusion 39.1%Readiness 75.7
International labor view
Workforce Structure by Major Economy
9 economies included. ISCO-08 major-group employment shares; source: ILOSTAT annual data, total employment, latest year within 3 years of 2025. National survey definitions, reference periods, and coverage differ — descriptive comparison only.
This section describes occupation composition only. No AI-exposure scores, wage rankings, or AI-impact claims are made. Displayed shares are normalized across the nine ISCO-08 major groups and sum to approximately 100%; each economy's coverage ratio separately discloses the share of total national employment those groups represent. Coverage is limited to economies with complete harmonized ISCO-08 data in the seed universe and does not represent all major economies.
Harmonized ISCO-08 major-group occupation shares across major economies
11 Managers
22 Professionals
33 Technicians and associate professionals
44 Clerical support workers
55 Service and sales workers
66 Skilled agricultural, forestry and fishery workers
77 Craft and related trades workers
88 Plant and machine operators, and assemblers
99 Elementary occupations
100% stacked horizontal bars showing each country's employment share across ISCO-08 major groups 1–9, ordered alphabetically by country name. This is not a ranked leaderboard.
Australia (2025): Managers 11.5%, Professionals 25.3%, Technicians and associate professionals 13.4%, Clerical support workers 8.5%, Service and sales workers 17.1%, Skilled agricultural, forestry and fishery workers 1.8%, Craft and related trades workers 9.6%, Plant and machine operators, and assemblers 5.8%, Elementary occupations 7.1%.
France (2025): Managers 7.6%, Professionals 25.9%, Technicians and associate professionals 17.9%, Clerical support workers 7.9%, Service and sales workers 14.3%, Skilled agricultural, forestry and fishery workers 2.5%, Craft and related trades workers 9.4%, Plant and machine operators, and assemblers 6.2%, Elementary occupations 8.4%.
Germany (2025): Managers 4.5%, Professionals 23.8%, Technicians and associate professionals 20.1%, Clerical support workers 12.4%, Service and sales workers 14.1%, Skilled agricultural, forestry and fishery workers 1.2%, Craft and related trades workers 10.9%, Plant and machine operators, and assemblers 5.6%, Elementary occupations 7.4%.
Italy (2025): Managers 4.0%, Professionals 16.2%, Technicians and associate professionals 17.5%, Clerical support workers 11.9%, Service and sales workers 17.8%, Skilled agricultural, forestry and fishery workers 2.2%, Craft and related trades workers 13.3%, Plant and machine operators, and assemblers 6.3%, Elementary occupations 10.7%.
Netherlands (2025): Managers 6.0%, Professionals 33.2%, Technicians and associate professionals 17.3%, Clerical support workers 8.2%, Service and sales workers 16.3%, Skilled agricultural, forestry and fishery workers 1.4%, Craft and related trades workers 6.4%, Plant and machine operators, and assemblers 3.9%, Elementary occupations 7.2%.
Republic of Korea (2025): Managers 1.4%, Professionals 23.1%, Technicians and associate professionals 17.9%, Clerical support workers 12.5%, Service and sales workers 8.7%, Skilled agricultural, forestry and fishery workers 4.7%, Craft and related trades workers 7.8%, Plant and machine operators, and assemblers 10.2%, Elementary occupations 13.6%.
Spain (2025): Managers 4.3%, Professionals 20.4%, Technicians and associate professionals 12.3%, Clerical support workers 9.8%, Service and sales workers 20.7%, Skilled agricultural, forestry and fishery workers 2.0%, Craft and related trades workers 10.9%, Plant and machine operators, and assemblers 7.4%, Elementary occupations 12.1%.
United Kingdom (2025): Managers 16.3%, Professionals 25.2%, Technicians and associate professionals 14.1%, Clerical support workers 8.9%, Service and sales workers 14.9%, Skilled agricultural, forestry and fishery workers 1.0%, Craft and related trades workers 6.1%, Plant and machine operators, and assemblers 4.3%, Elementary occupations 9.1%.
United States of America (2025): Managers 11.7%, Professionals 23.0%, Technicians and associate professionals 17.8%, Clerical support workers 8.3%, Service and sales workers 14.7%, Skilled agricultural, forestry and fishery workers 0.4%, Craft and related trades workers 8.1%, Plant and machine operators, and assemblers 5.6%, Elementary occupations 10.4%.
Full data table
ISCO-08 employment shares by country
Country
Year
1
2
3
4
5
6
7
8
9
Coverage
Status
Australia
2025
11.5%
25.3%
13.4%
8.5%
17.1%
1.8%
9.6%
5.8%
7.1%
100.0%
—
France
2025
7.6%
25.9%
17.9%
7.9%
14.3%
2.5%
9.4%
6.2%
8.4%
98.3%
—
Germany
2025
4.5%
23.8%
20.1%
12.4%
14.1%
1.2%
10.9%
5.6%
7.4%
99.1%
—
Italy
2025
4.0%
16.2%
17.5%
11.9%
17.8%
2.2%
13.3%
6.3%
10.7%
99.0%
—
Netherlands
2025
6.0%
33.2%
17.3%
8.2%
16.3%
1.4%
6.4%
3.9%
7.2%
99.2%
—
Republic of Korea
2025
1.4%
23.1%
17.9%
12.5%
8.7%
4.7%
7.8%
10.2%
13.6%
99.6%
—
Spain
2025
4.3%
20.4%
12.3%
9.8%
20.7%
2.0%
10.9%
7.4%
12.1%
99.6%
—
United Kingdom
2025
16.3%
25.2%
14.1%
8.9%
14.9%
1.0%
6.1%
4.3%
9.1%
99.8%
—
United States of America
2025
11.7%
23.0%
17.8%
8.3%
14.7%
0.4%
8.1%
5.6%
10.4%
100.0%
Break in series
ISCO-08 employment shares (fraction of total national employment) for 9 included countries. Values reflect each country's latest year within a 3-year window of 2025.
Not included in comparable set: Canada, Japan. Canada: No ISCO-08 data within 3 years of dataset latest year (2025); Japan: Insufficient ISCO-08 groups in all qualifying years: only 7 of 9 present at latest year 2025 (missing groups: 3,7).
Select a country bar above to see its detailed ISCO-08 occupation-mix breakdown.
Source: ILOSTAT — Employment by sex and occupation (EMP_TEMP_SEX_OCU_NB_A).ilostat.ilo.org · CC BY 4.0 · International Labour Organization (ILO)
9 economies passed the minimum coverage filter (all 9 of 9 ISCO-08 groups present, ≥ 98% of national employment covered, within 3 years of dataset latest year). Coverage is verified, not complete.
Not included in comparable set: Canada: No ISCO-08 data within 3 years of dataset latest year (2025); Japan: Insufficient ISCO-08 groups in all qualifying years: only 7 of 9 present at latest year 2025 (missing groups: 3,7).
For U.S.-only AI occupation-exposure analysis, see the Analysis page — a separate U.S.-only view using different methodology and data. It is not merged with the international occupation-mix data shown here.
Countries with the largest GenAI diffusion gains, H1 2025 → Q2 2026. Based on Microsoft’s AI Economic Impact Index (Western telemetry — may undercount domestic apps in some markets). Full source details →
South Korea
25.9% → 40.6%+14.7pp
United Arab Emirates
59.4% → 73.3%+13.9pp
France
40.9% → 49.6%+8.7pp
Ireland
41.7% → 49.9%+8.2pp
Japan
16.7% → 24.7%+8.0pp
Microsoft AIEI · H1 2025 → Q2 2026 · % working-age population using generative AI across 147 economies. Western telemetry — see sources for caveats.
Microsoft AIEI · MIT License
Consumer GenAI Diffusion — Top Economies
Top 10 economies ranked by Q2 2026 share of working-age population using a generative AI product. H1 2025, H2 2025, Q1 2026, and Q2 2026 values shown for trend context. Ranked by Q2 2026 level, descending — not a fastest-growth ranking.
Usage ≠ capability, workplace adoption, productivity, or labor-market impact.
Consumer GenAI diffusion — top 10 economies by Q2 2026 share (Microsoft AIEI, MIT)
Economy
H1 2025 (%)
H2 2025 (%)
Q1 2026 (%)
Q2 2026 (%)
Change (Q2 2026−H1 2025, pp)
United Arab EmiratesARE
59.4%
64.0%
70.1%
73.3%
+13.9pp
SingaporeSGP
58.6%
60.9%
63.4%
64.3%
+5.7pp
IrelandIRL
41.7%
44.6%
48.4%
49.9%
+8.2pp
FranceFRA
40.9%
44.0%
47.8%
49.6%
+8.7pp
NorwayNOR
45.3%
46.4%
48.6%
49.4%
+4.1pp
SpainESP
39.7%
41.8%
44.2%
45.1%
+5.4pp
New ZealandNZL
37.6%
40.5%
43.0%
44.1%
+6.5pp
United KingdomGBR
36.4%
38.9%
42.2%
43.7%
+7.3pp
The NetherlandsNLD
36.3%
38.9%
42.1%
43.5%
+7.2pp
QatarQAT
35.7%
38.3%
41.8%
43.5%
+7.8pp
Metric: % of working-age population who reported using a generative AI product in each survey period (Microsoft AI Economic Impact & Insights). Usage ≠ capability, workplace adoption, productivity, or labor-market impact. Four survey periods (H1 2025, H2 2025, Q1 2026, Q2 2026) is a short window; caution on trend extrapolation. This absolute-share top 10 reflects economies with high Microsoft product penetration and digital-access infrastructure; it is not a representative sample of global AI diffusion. Digital-access gaps and Microsoft product reach independently affect which economies appear in this ranking. Western telemetry may undercount domestic AI apps (e.g. Doubao, Kimi) in China and other markets. Source: Microsoft AI Diffusion Report (MIT). Not merged with Claude usage index, Indeed job demand, Anthropic indices, or IMF metrics.
World Map — AI Usage Index
Per-capita AI usage index normalised against working-age population. Darker / higher = more AI usage relative to population size.
Australia: usage index 6.40, global share 2.65%
Singapore: usage index 5.81, global share 0.62%
Switzerland: usage index 5.02, global share 0.70%
Luxembourg: usage index 4.85, global share 0.05%
New Zealand: usage index 4.84, global share 0.39%
Canada: usage index 4.13, global share 2.63%
Norway: usage index 4.04, global share 0.35%
Iceland: usage index 4.02, global share 0.02%
Malta: usage index 4.00, global share 0.04%
France: usage index 3.97, global share 3.95%
The Netherlands: usage index 3.90, global share 1.07%
United States: usage index 3.87, global share 20.16%
Belgium: usage index 3.80, global share 0.68%
South Korea: usage index 3.78, global share 3.24%
Denmark: usage index 3.57, global share 0.32%
United Kingdom: usage index 3.35, global share 3.47%
Portugal: usage index 3.32, global share 0.53%
Ireland: usage index 3.23, global share 0.27%
Estonia: usage index 3.21, global share 0.07%
Cyprus: usage index 3.08, global share 0.07%
Top Countries by AI Adoption
Ranked by usage index (per-capita Claude.ai usage, normalised). Countries with zero recorded usage or unreported Claude.ai metrics are excluded. Click any row or use the selector to view the full metric set.
Methodology
Usage index = observed Claude.ai interactions per working-age capita, normalised across all countries. Source: Anthropic Economic Index, June 2026 release (latest reported month per country, Apr–May 2026; countries below the release’s reporting threshold have no usage value, plus a supplemental China row using World Bank 2024 GDP and working-age population). GDP data comes from World Bank / IMF fields bundled in the Anthropic dataset, with China GDP-per-worker sourced directly from World Bank. Countries with zero recorded interactions are excluded from ranked lists but remain in the dataset; countries with unreported Claude.ai usage metrics do not rank.
For full details on data provenance and licensing, see the Data & Sources page.