in/out

Explore the evidence / Organisations

From adoption to integration.

Enterprise use across the EU and UK, with distinct survey populations, sector differences and organisational scaling.

Find a chart, dataset, or source

Each chart keeps its own population and observation period. Wide charts can be scrolled horizontally; open the values beneath any chart for the underlying observations.

01 / Where firms use AI

Adoption differs across European economies.

In 2025, 19.95% of survey-covered EU enterprises used AI. Denmark reached 42.03%, Finland 37.82% and Romania 5.21%. These are shares of firms in eligible activities, not shares of people.

Where European firms have adopted AI

27 EU member states · 2025 · enterprises with 10+ persons employed

Enterprise adoption: European Union, 2025Eurostat 2025 data for 27 EU member states. Percentage of enterprises with 10 or more persons employed in survey-covered industries using at least one AI technology. Common 0–50% scale. Grey countries are outside this panel’s coverage. Malta and Luxembourg use schematic markers. This is broad AI, not only generative AI.Tanzania: No published observationWestern Sahara: No published observationCanada: No published observationUnited States: No published observationKazakhstan: No published observationUzbekistan: No published observationCongo (DRC): No published observationSomalia: No published observationKenya: No published observationSudan: No published observationChad: No published observationRussia: No published observationNorway: No published observationGreenland: No published observationFrance: 18.2%Senegal: No published observationMali: No published observationMauritania: No published observationBenin: No published observationNiger: No published observationNigeria: No published observationCameroon: No published observationTogo: No published observationGhana: No published observationCote D'Ivoire: No published observationGuinea: No published observationGuinea-Bissau: No published observationLiberia: No published observationSierra Leone: No published observationBurkina Faso: No published observationCentral African Republic: No published observationCongo: No published observationGabon: No published observationEquatorial Guinea: No published observationIsrael: No published observationLebanon: No published observationPalestine: No published observationGambia: No published observationTunisia: No published observationAlgeria: No published observationJordan: No published observationUnited Arab Emirates: No published observationQatar: No published observationKuwait: No published observationIraq: No published observationOman: No published observationMongolia: No published observationIndia: No published observationNepal: No published observationPakistan: No published observationAfghanistan: No published observationTajikistan: No published observationKyrgyzstan: No published observationTurkmenistan: No published observationIran: No published observationSyria: No published observationArmenia: No published observationSweden: 35.0%Belarus: No published observationUkraine: No published observationPoland: 8.4%Austria: 30.0%Hungary: 10.4%Moldova: No published observationRomania: 5.2%Lithuania: 21.3%Latvia: 12.2%Estonia: 23.4%Germany: 26.0%Bulgaria: 8.6%Greece: 8.9%Türkiye: No published observationAlbania: No published observationCroatia: 15.2%Switzerland: No published observationLuxembourg: 33.6%Belgium: 34.5%Netherlands: 33.2%Portugal: 11.5%Spain: 20.3%Ireland: 19.6%China: No published observationItaly: 16.4%Denmark: 42.0%United Kingdom: No published observationIceland: No published observationAzerbaijan: No published observationGeorgia: No published observationSlovenia: 21.6%Finland: 37.8%Slovakia: 18.0%Czech Republic: 17.6%Eritrea: No published observationYemen: No published observationSaudi Arabia: No published observationTurkish Republic of Northern Cyprus: No published observationCyprus: 9.3%Morocco: No published observationEgypt: No published observationLibya: No published observationEthiopia: No published observationDjibouti: No published observationSomaliland: No published observationUganda: No published observationRwanda: No published observationBosnia And Herzegovina: No published observationNorth Macedonia: No published observationSerbia: No published observationMontenegro: No published observationKosovo: No published observationSouth Sudan: No published observationLuxembourg: 33.6%; schematic markerMalta: 21.5%; schematic markerDenmark 42.0%Finland 37.8%Romania 5.2%Italy 16.4%Luxembourg 33.6%Malta 21.5%0%10%20%30%40%50%No published observation / outside coverage

Scroll the map sideways on smaller screens. Exact values are available in the linked data tables.

Eurostat, isoc_eb_ai, snapshot updated 15 June 2026. Broad AI use in survey-covered activities; finance and several public-service sectors are excluded. Denmark 42.03%, Finland 37.82%, Romania 5.21%; EU aggregate 19.95%. Netherlands 2025 carries a break in time series. Exact country evidence. Natural Earth geometry, public domain. Small circles are schematic country markers; area is not weighted by population.

Regional evidence / OECD · 2025 data

Business adoption has a geography too.

In the selected 2025 data, the simple average regional share of enterprises using AI was 11.43% across 43 Central and Eastern European regions, compared with 26.10% across 48 other covered regions. These are unweighted means of regional rates across 12 countries, not enterprise-weighted adoption rates for Europe.

Simple means of regional enterprise AI adoption rates2025: Central and Eastern Europe, 43 covered regions, 11.43 percent; other covered regions, 48 regions, 26.10 percent. Selected 12-country coverage. Derived regional averages, not enterprise-weighted Europe-wide rates.Central & Eastern · 43 regions11.43%Other covered · 48 regions26.10%0100%
Derived from OECD Figure 1.9. Each region receives equal weight. Enterprises with at least 10 persons employed in covered nonfinancial sectors; broad AI use. Publication: 4 November 2026.

OECD analysis of existing Eurostat isoc_r_eb_ain2 data, not an independent survey. Selected 2025 regional coverage in 12 countries; not Europe-wide. Regional achieved sample counts and confidence intervals are not supplied in the workbook. Employment exposure in column D is excluded: it has a different denominator.

Covered countries: Austria, Belgium, Croatia, Czechia, Denmark, Hungary, Lithuania, Norway, Poland, Slovak Republic, Spain, Sweden.

All 91 regional values and populations
Enterprises using AI, 2025 · selected OECD regional coverage · percent
CountryRegionCodeAI useOECD grouping
AustriaBurgenlandAT1118.33%North, South & West Europe
AustriaCarinthiaAT2123.63%North, South & West Europe
AustriaLower AustriaAT1221.81%North, South & West Europe
AustriaSalzburgAT3231.45%North, South & West Europe
AustriaStyriaAT2225.55%North, South & West Europe
AustriaTyrolAT3330.28%North, South & West Europe
AustriaUpper AustriaAT3132.93%North, South & West Europe
AustriaViennaAT1337.36%North, South & West Europe
AustriaVorarlbergAT3436.42%North, South & West Europe
BelgiumBrussels Capital RegionBE142.32%North, South & West Europe
BelgiumFlemish RegionBE234.84%North, South & West Europe
BelgiumWalloon RegionBE330.20%North, South & West Europe
CroatiaAdriatic CroatiaHR0310.98%East Europe
CroatiaCity of ZagrebHR0525.44%East Europe
CroatiaNorthern CroatiaHR0611.98%East Europe
CroatiaPannonian CroatiaHR0210.51%East Europe
CzechiaCentral Bohemian RegionCZ0213.47%East Europe
CzechiaCentral MoraviaCZ0715.57%East Europe
CzechiaMoravia-SilesiaCZ0815.72%East Europe
CzechiaNortheastCZ0517.81%East Europe
CzechiaNorthwestCZ049.47%East Europe
CzechiaPragueCZ0125.50%East Europe
CzechiaSoutheastCZ0615.49%East Europe
CzechiaSouthwestCZ0314.06%East Europe
DenmarkCentral JutlandDK0450.82%North, South & West Europe
DenmarkCopenhagen regionDK0133.05%North, South & West Europe
DenmarkNorthern JutlandDK0532.13%North, South & West Europe
DenmarkSouthern DenmarkDK0334.68%North, South & West Europe
DenmarkZealandDK0242.73%North, South & West Europe
HungaryBudapestHU1117.66%East Europe
HungaryCentral TransdanubiaHU217.72%East Europe
HungaryNorthern Great PlainHU325.48%East Europe
HungaryNorthern HungaryHU316.26%East Europe
HungaryPestHU127.40%East Europe
HungarySouthern Great PlainHU337.07%East Europe
HungarySouthern TransdanubiaHU237.98%East Europe
HungaryWestern TransdanubiaHU226.11%East Europe
LithuaniaCentral and Western LithuaniaLT0217.98%East Europe
LithuaniaVilnius RegionLT0126.84%East Europe
NorwayAgder and Sør-ØstlandetNO0924.22%North, South & West Europe
NorwayInnlandetNO0223.64%North, South & West Europe
NorwayNorthern NorwayNO0720.38%North, South & West Europe
NorwayOslo and VikenNO0835.49%North, South & West Europe
NorwayTrøndelagNO0627.18%North, South & West Europe
NorwayWestern NorwayNO0A25.56%North, South & West Europe
PolandGreater PolandPL417.50%East Europe
PolandKuyavian-PomeraniaPL616.36%East Europe
PolandLesser PolandPL218.86%East Europe
PolandLodzkiePL717.57%East Europe
PolandLower SilesiaPL519.68%East Europe
PolandLublin ProvincePL815.86%East Europe
PolandLubuszPL435.58%East Europe
PolandMazowiecki regionPL922.94%East Europe
PolandOpole regionPL527.40%East Europe
PolandPodkarpaciaPL827.02%East Europe
PolandPodlaskiePL845.56%East Europe
PolandPomeraniaPL639.24%East Europe
PolandSilesiaPL228.05%East Europe
PolandSwietokrzyskiePL725.31%East Europe
PolandWarmian-MasuriaPL624.53%East Europe
PolandWarsawPL9115.99%East Europe
PolandWest PomeraniaPL427.92%East Europe
Slovak RepublicBratislava RegionSK0128.36%East Europe
Slovak RepublicCentral SlovakiaSK0314.00%East Europe
Slovak RepublicEast SlovakiaSK0413.36%East Europe
Slovak RepublicWest SlovakiaSK0214.00%East Europe
SpainAndalusiaES6115.83%North, South & West Europe
SpainAragonES2412.68%North, South & West Europe
SpainAsturiasES1215.08%North, South & West Europe
SpainBalearic IslandsES5311.87%North, South & West Europe
SpainBasque CountryES2122.42%North, South & West Europe
SpainCanary IslandsES7015.85%North, South & West Europe
SpainCantabriaES1310.99%North, South & West Europe
SpainCastile and LeónES4117.13%North, South & West Europe
SpainCastile-La ManchaES4218.93%North, South & West Europe
SpainCataloniaES5124.04%North, South & West Europe
SpainExtremaduraES4314.88%North, South & West Europe
SpainGaliciaES1117.06%North, South & West Europe
SpainLa RiojaES2314.61%North, South & West Europe
SpainMadridES3027.42%North, South & West Europe
SpainMurciaES6215.09%North, South & West Europe
SpainNavarraES2220.35%North, South & West Europe
SpainValenciaES5219.82%North, South & West Europe
SwedenCentral NorrlandSE3218.43%North, South & West Europe
SwedenEast Middle SwedenSE1229.23%North, South & West Europe
SwedenNorth Middle SwedenSE3122.09%North, South & West Europe
SwedenSmåland with IslandsSE2129.53%North, South & West Europe
SwedenSouth SwedenSE2231.32%North, South & West Europe
SwedenStockholmSE1144.77%North, South & West Europe
SwedenUpper NorrlandSE3331.51%North, South & West Europe
SwedenWest SwedenSE2337.08%North, South & West Europe

Source and methods · Regional values · CSV · Data and provenance · JSON · OECD workbook ↗

UK businesses · separate survey

The size gradient extends beyond the EU.

32% of eligible UK businesses report using AI, including pilots. Use rises from 27% among micro businesses to 52% among large businesses. This survey measures broad AI, not only generative AI; its population differs from the EU series above.

UK AI use by business sizeMicro 27%, small 34%, medium 47%, large 52%. Zero to one hundred percent scale; UK survey only.Micro · 5–9 employees27%Small · 10–49 employees34%Medium · 50–249 employees47%Large · 250+ employees52%0%25%50%75%100%
UK current AI use, including pilots · 7 January–21 April 2026 · weighted percentages
Business sizeAI useUnweighted base
All eligible businesses32%3,503
Micro · 5–9 employees27%1,347
Small · 10–49 employees34%1,579
Medium · 50–249 employees47%447
Large · 250+ employees52%130
UK private-sector businesses with at least five employees in SIC A–N and R–S. 3,503 telephone interviews with technology-responsible respondents; Market Location commercial sampling frame; weighted by business size and sector; 5% response rate. Table 27 · Sources and methods.

Capability remains a separate constraint. Among nonusers planning adoption, 78% cite limited AI skills, expertise or knowledge (unweighted n=236; multiple responses allowed). This is not a share of all businesses or current adopters. Table 61.

Broad AI, including pilots, not only generative AI. Selected sectors and a five-employee threshold differ from Eurostat; no direct UK–EU ranking. Self-reports, not a census or causal productivity evidence. Low response creates nonresponse risk; weighting does not establish its absence. The report compares 32% with 16% in the prior wave. These are repeated cross-sections, not a panel; questionnaire and achieved samples changed.

Download UK observations and denominators · CSV · Evidence and methods · JSON

02 / Which activities are represented

Read sector growth with its coverage.

The sector series describes eligible activities. Healthcare services are excluded from the healthcare and pharmaceuticals estimate; farming, fishing and aquaculture are excluded from agri-food. A high adoption rate in a narrow slice does not describe the whole sector.

Sector growth needs a coverage lens

EU eligible firms using AI. Open dots: 2021; filled dots: 2025. Both endpoints use a 0–50% scale.

AI adoption across ten EU strategic sectors, 2021 and 2025Open dots show 2021; solid dots show 2025. Axis zero to fifty percent. Survey coverage is shown separately as a percentage of sector value added, not a confidence interval.2021 → 2025Coverage¹Creative sectors & media13 → 38%85%Energy18 → 34%100%Electronic communications12 → 30%100%Healthcare / pharmaceuticals7 → 23%11%Defence, security & space9 → 20%100%Climate & environment7 → 16%90%Manufacturing & construction7 → 16%100%Agri-food5 → 13%8%Tourism3 → 13%92%Mobility, transport and autom…6 → 12%100%01020304050
¹ Coverage is the share of sector gross value added represented by the survey’s activities. Healthcare services are excluded from the healthcare/pharma estimate; agriculture and fishing are excluded from agri-food. JRC, Figure 3, p. 11; coverage Table 2, p. 34.
Exact sector values and exclusions
Sector20212025CoverageExcluded activities
Cultural and creative sectors, and media13%38%85%Arts, cultural establishments and cultural education
Energy18%34%100%None listed
Electronic communications12%30%100%None listed
Healthcare and pharmaceuticals (covered activities)7%23%11%Healthcare services including public hospitals and healthcare centres
Defence, security and space (private activities)9%20%100%Public-sector defence activities
Climate and environment7%16%90%Forest management; some other activities lack data
Manufacturing, engineering and construction7%16%100%None listed
Agri-food (covered activities)5%13%8%Farming, fishing and aquaculture
Tourism3%13%92%Sports, games, amusement parks and spa activities
Mobility, transport and automotive6%12%100%None listed

OECD / Technology combinations

Using AI alongside cloud and analytics.

Around 12% is the unweighted mean across 28 covered OECD countries for enterprises with at least ten employees using all three technologies. It measures co-use; it does not establish successful workflow integration or returns.

Published mean and 28 country rates (displayed to four decimals)
OECD Figure 3.4 — published mean and 28 covered country rates
CountryReference periodAI + cloud + analytics (%)
OECD2025 or most recent (2021–2025)12.2162
Australia (2022)2021–20221.4000
Austria202511.1124
Belgium202522.7793
Canada (2023)20231.8000
Czech Republic202511.0392
Denmark202529.9112
Estonia202516.1476
Finland202525.4826
France20258.3080
Germany202512.9291
Greece20254.6703
Hungary20255.7149
Ireland202513.4635
Italy202510.1717
Japan (2023)2021–20232.9417
Korea (2024)202420.8704
Latvia20258.1669
Lithuania202514.6097
Luxembourg202514.2910
Netherlands202523.7163
Norway202516.1432
Poland20255.2906
Portugal20257.2088
Slovak Republic20258.6340
Slovenia202510.9384
Spain202510.6999
Sweden202520.8595
Türkiye20252.7536

Mostly 2025, with older observations and differing national definitions. Definitions, coverage and calculation.

OECD chapter / Figure 3.4 ↗ · Original workbook ↗ · Workbook snapshot · 29 observations · CSV · Data & provenance · JSON

United States / Census BTOS

Two business measures, separate questions.

US BTOS — 5–18 October 2026
MeasureShare of businessesStandard error (pp)
AI use in any business function24.2%0.61
Business reports employees using generative AI for work26.8%0.56

Both figures are percentages of covered nonfarm employer businesses, not employees. Reference period 5–18 October; collection 19 October–1 November; released 5 November 2026. The questions overlap. Wording changed to SI in this collection, so no trend is inferred or direct UK/EU ranking made. Sampling and comparability.

5 November release ↗ · Survey notes ↗ · Methodology ↗ · National workbook snapshot · Two observations · CSV · Data & provenance · JSON

03 / From use to deployment

Size, scaling and returns answer different questions.

McKinsey defines large organisations by revenue above $1 billion. Its agent-scaling comparison is separate from the questions about personal productivity and organisational profit; the latter are not a conversion funnel.

Larger firms pull ahead with agents

Respondents reporting agent scaling in one or more functions; 0–100%.

Scaling agents: large and smaller organisationsLarge organisations: 27% in 2025, 40% in 2026. Smaller organisations: 22% in both years. Zero to one hundred percent scale.Large organisations27 → 40%Smaller organisations22 → 22%0255075100%20252026
Exact values and populations
The state of AI in 2026: On the road to ROI. Respondents reporting agent scaling in one or more functions; 0–100%.
Measure / groupShareObservation periodPopulation
Large organisations27%2025Respondents at organisations with annual revenue >$1billion
Large organisations40%2026Respondents at organisations with annual revenue >$1billion
Smaller organisations22%2025Respondents at smaller organisations; large defined as annual revenue >$1billion
Smaller organisations22%2026Respondents at smaller organisations; large defined as annual revenue >$1billion
Large means annual revenue above $1 billion. McKinsey, Exhibit 2.

Personal gains and organisational returns

Two different questions; separate responses, not a conversion funnel.

Reported productivity and profit effectsImproved own productivity: 80 percent; Positive organisational EBIT contribution: 37 percent. Common scale zero to one hundred percent.Improved own productivity80%Positive organisational EBITcontribution37%0255075100%
Exact values and populations
The state of AI in 2026: On the road to ROI. Two different questions; separate responses, not a conversion funnel.
Measure / groupShareObservation periodPopulation
Improved own productivity80%2026Survey respondents describing their own productivity
Positive organisational EBIT contribution37%2026Survey respondents attributing at least some positive enterprise EBIT contribution to AI
McKinsey, individual and enterprise effects.

04 / The annual comparison

Deployment expanded across functions.

Enterprise-wide scaling rose from 38% to 44%; use in three or more functions rose from 51% to 56%. These annual samples describe expanding deployment, not proof of attributable returns.

Exact values and populations
McKinsey annual comparisons reported in the 2026 survey. Different annual samples.
Measure / groupShareObservation periodPopulation
Enterprise-wide AI scaling38%2025Survey respondents reporting on their organisations
Enterprise-wide AI scaling44%2026Survey respondents reporting on their organisations
AI in three or more functions51%2025Survey respondents reporting on their organisations
AI in three or more functions56%2026Survey respondents reporting on their organisations

McKinsey, 25 August 2026. Fieldwork: 4 May–8 June 2026; 1,719 respondents in 97 nations, weighted by national GDP. Self-reports; not a firm census or causal evaluation.

05 / Organisational redesign

Deployment does not describe how work has changed.

In Kyndryl’s survey, 84% agreed AI was deployed beyond pilots, while 27% completely agreed roles had been redesigned around AI and 29% completely agreed formal change management was fully implemented. These statements use different response thresholds.

Deployment and organisational redesign. Kyndryl 2026–27 Readiness Report. Senior-leader responses, n=3,700; fieldwork 10 July–2 September 2026. Different response thresholds; not a subtraction-based integration gap.
AI deployed beyond pilot projects · Agree84%
Roles formally redesigned or reorganised around AI · Completely agree27%
Formal AI change-management programmes fully implemented · Completely agree29%
Data
Kyndryl 2026–27 Readiness Report. Senior-leader responses, n=3,700; fieldwork 10 July–2 September 2026. Different response thresholds; not a subtraction-based integration gap.
StatementShareResponse thresholdSource
AI deployed beyond pilot projects84%AgreeQ8; p.9
Roles formally redesigned or reorganised around AI27%Completely agreeQ6; p.25
Formal AI change-management programmes fully implemented29%Completely agreeQ6; p.25

Leadership self-reports, not a representative national firm-adoption estimate. Half the respondents represented companies with at least $1 billion in revenue. The report does not disclose recruitment frame, response rate, weighting or continuity of respondents across years. Different agreement thresholds mean the percentages cannot be subtracted to calculate an integration gap or interpreted as stages in a funnel. Methods · Primary report · CSV

06 / Spending intensity

Spending is a different dimension of diffusion.

Menlo’s survey reports AI spending at 4.7 times its prior-year level on average, versus a median of 2.2 times. The distance between them suggests spending growth is uneven within the sample. It does not measure how widely AI is adopted or whether it improves productivity.

Spending growth is uneven. Reported 2026 AI spending relative to 2025; surveyed US enterprises. Multiples, not percentages or adoption rates.
Measure0×5×Multiple
Average4.7×
Median2.2×
Data
Menlo 2026 survey. Fieldwork: 8–26 September 2026. Reported spending relative to 2025.
MeasureMultiplePopulation
Average4.7×Surveyed US enterprises
Median2.2×Surveyed US enterprises

Published 3 November 2026. Morning Consult surveyed nearly 400 US enterprise decision-makers on 8–26 September 2026. Menlo does not disclose the recruitment frame, response rate or representative weighting. Menlo invests in companies discussed in the report. Methods and market-estimate caveats · Primary source ↗ · Report PDF ↗ · CSV · Data and provenance · JSON

Reports, methods and downloadable observations →