ai statistics·Data & Research

    AI Adoption by Country 2026: Complete Rankings

    Enterprise AI adoption rates across 30+ countries, drawn from Eurostat (2025 reference year), OECD, UK DSIT and US Census — with methodology, sources, and what's driving the leaders.

    AI adoption by country measures the percentage of enterprises in a given country that have used at least one artificial intelligence technology in production. In the latest Eurostat release (2025 reference year, published December 2025), reported adoption rates range from 5.2% in Romania to 42.0% in Denmark, with an EU27 average of 20.0% and an OECD firm average of 20.2%.

    Linus IngemarssonCo-Founder, Alice Labs
    (Updated: )
    10 min read
    Last reviewed:
    Quick Answer
    Denmark leads AI adoption in 2026 at 42.0% of enterprises, followed by Finland (37.8%), Sweden (35.0%) and Belgium (34.5%). The EU27 average is 20.0% (Eurostat 2025 reference year). OECD firm adoption averages 20.2%.
    20.0%

    EU27 enterprise AI adoption

    +6.5 pp vs 2024

    Eurostat 2025 reference year

    42.0%

    Denmark — #1 globally

    Highest EU27 adoption

    Eurostat ISOC_EB_AI 2025

    20.2%

    OECD firm AI adoption

    Doubled since 2023

    OECD AI in Business 2025

    35 pp

    Large vs small firm gap (OECD)

    Persistent digital divide

    OECD 2025

    What you'll learn

    • How enterprise AI adoption is measured and why numbers vary across sources
    • The 2026 ranking of EU countries by enterprise AI adoption rate (Eurostat 2025 data)
    • What separates AI leaders (Denmark, Finland, Sweden) from laggards (Romania, Bulgaria)
    • How the US (17.3%), UK (16%) and OECD average (20.2%) compare to the EU baseline
    • Methodology, survey scope, and citable sources for journalists and researchers

    Key Takeaways

    • EU27 enterprise AI adoption reached 20.0% in the 2025 reference year — up from 13.5% in 2024, according to Eurostat's ICT Enterprise Survey.
    • Denmark (42.0%), Finland (37.8%), Sweden (35.0%) and Belgium (34.5%) lead European adoption; Romania (5.2%), Bulgaria (8.5%) and Poland (8.4%) trail.
    • OECD firm-level AI adoption averaged 20.2% in 2025, more than doubling from 8.7% in 2023 — and roughly tripling since 2022.
    • US BTOS data shows 17.3% of businesses use AI for any function (Nov 2025); UK DSIT reports 16% of UK businesses with 5+ employees.
    • A 35-point gap separates large firms (~52% OECD average) from small firms (~17%) — the persistent small-enterprise digital divide.

    How AI Adoption Is Measured

    AI adoption by country is measured through harmonised national ICT surveys — primarily Eurostat's ICT Enterprise Survey in Europe, the OECD AI in Business benchmark, UK DSIT's AI Adoption Research, and the US Census Bureau's Business Trends and Outlook Survey (BTOS).

    "AI adoption by country" sounds simple, but the number you quote depends heavily on the source. Eurostat asks enterprises with 10+ employees whether they used at least one of six AI technologies in the previous 12 months. The OECD compiles firm-level data across its 38 member countries. UK DSIT surveys businesses with 5+ employees. US Census BTOS asks "any business" across multiple AI definitions.

    For cross-country comparability in Europe, the Eurostat ICT Enterprise Survey remains the gold standard: identical questions, same reference period, all 27 member states. The 2025 reference year data (published December 2025) is the latest harmonised release and what this article uses as its spine.

    European Rankings (Eurostat 2025 Reference Year)

    Among EU27 member states, Denmark leads at 42.0% enterprise adoption, followed by Finland (37.8%), Sweden (35.0%), Belgium (34.5%) and the Netherlands (33.2%). Romania (5.2%), Poland (8.4%) and Bulgaria (8.5%) trail the ranking.

    The table below ranks EU27 countries by share of enterprises (10+ employees) using at least one AI technology, from Eurostat's December 2025 release (2025 reference year). Sort any column to re-rank. Country-level breakdowns for Germany, France, Italy, Spain and Ireland for 2025 have not yet been published individually in Eurostat news releases — for those, consult the interactive databrowser.

    Enterprise AI adoption by EU country (10+ employees, 2025 reference year). Source: Eurostat ICT Enterprise Survey, release DDN-20251211-2.
    Rank Country Adoption rate vs EU27 avg
    1 Denmark 42.0% +22.0 pp
    2 Finland 37.8% +17.8 pp
    3 Sweden 35.0% +15.0 pp
    4 Belgium 34.5% +14.5 pp
    5 Netherlands 33.2% +13.2 pp
    EU27 average 20.0%
    25 Bulgaria 8.5% −11.5 pp
    26 Poland 8.4% −11.6 pp
    27 Romania 5.2% −14.8 pp

    Source: Eurostat ISOC_EB_AI (2025 reference year)

    Global Comparison: US, UK, OECD, Stanford HAI

    Outside the EU, OECD firm-level adoption averaged 20.2% in 2025 (up from 8.7% in 2023). UK DSIT reports 16% of businesses with 5+ employees use AI. US Census BTOS shows 17.3% of businesses use AI for any function (November 2025 wave).

    The OECD's AI in Business 2025 update places firm-level adoption at an average of 20.2% across OECD economies, up from 14.2% in 2024 and 8.7% in 2023. That trajectory — a doubling in two years — is the headline most enterprise leaders should internalise.

    Beneath the average, OECD data shows sectoral variation: the ICT sector leads at 57.3% adoption, while large firms (52.0%) adopt at roughly 3x the rate of small firms (17.4%).

    National surveys vary in scope and methodology. Below are the most reliable harmonised readings for major non-EU markets:

    20.2%

    OECD firm AI adoption

    +11.5 pp since 2023

    OECD 2025

    17.3%

    US business AI use (any function)

    vs 10% for production use

    US Census BTOS, Nov 2025

    16.0%

    UK enterprise AI adoption (5+ employees)

    Fieldwork Feb–May 2025

    UK DSIT AI Adoption Research 2025

    Stanford HAI: Fortune 500 and Global Enterprise Benchmarks

    Stanford HAI's AI Index 2025 reports 78% of Fortune 500 companies have active generative AI initiatives, 67% of enterprises globally report active AI initiatives, and the average enterprise now uses 5.3 different GenAI applications.

    Stanford HAI's AI Index Report 2025 uses a broader definition than Eurostat or OECD — measuring initiatives and pilots, not production deployment. That difference explains the higher headline numbers:

    • 78% of Fortune 500 companies report active generative AI initiatives
    • 67% of enterprises globally report at least one active AI initiative
    • Average enterprise uses 5.3 different GenAI applications across workflows

    Turn these numbers into strategy

    Book a 30-minute benchmark call with our AI strategists — we'll show you how your industry and country rank against the leaders, and where the closest wins are.

    Book benchmark call

    What Drives the Leaders

    Top-adopting countries (Denmark, Finland, Sweden, Belgium, Netherlands) share four traits: strong digital infrastructure, coordinated national AI strategies with public funding, large STEM talent cohorts, and early private-sector investment in cloud AI platforms.

    Adoption leadership rarely comes from a single factor. Across the five leading EU countries, four structural advantages consistently appear:

    • Digital infrastructure: Nordic and Benelux broadband penetration exceeds 95% and cloud-first IT strategies are common.
    • National strategies with budget: AI Sweden, Denmark's Digital Growth Plan, Finland's AI Programme, and Belgium's AI 4 Belgium all attach funding to ambitions.
    • Talent density: All five leaders rank in the top 15 globally for STEM graduates per capita.
    • Private-sector early adoption: Large banks, industrial firms and public utilities piloted AI in 2018–2022, normalising enterprise use earlier than peers.

    Methodology & Limitations

    Rankings use Eurostat's ICT Enterprise Survey (EU27, 10+ employees, 2025 reference year) as the primary source, supplemented by OECD AI in Business (2025), UK DSIT (2025), US Census BTOS (Nov 2025), and Stanford HAI AI Index (2025). All figures are for firms or enterprises, not population-level AI use.

    Scope: Enterprises/businesses with 10 or more employees (Eurostat, OECD); 5+ employees (UK DSIT); all sizes (US Census BTOS, Stanford HAI).

    AI definition (Eurostat): Deployment of one or more of six AI technology groups — text mining, natural language processing (NLP/chat), speech recognition, image recognition/computer vision, machine learning, or AI-based autonomous agents.

    Reference period: For Eurostat, the 12 months preceding the survey. The December 2025 release reflects activity during 2025.

    What this data does not measure: Consumer AI use, depth of deployment (one pilot vs. enterprise-wide production), or economic impact. For deeper Sweden analysis, see our Global AI Adoption Index 2026 and State of AI in Sweden 2026.

    Written by

    Linus Ingemarsson

    Co-Founder, Alice Labs

    Linus is co-founder of Alice Labs, an AI consultancy that has delivered 50+ enterprise AI implementations across Sweden and Europe. He writes on AI adoption, strategy, and measurement using public statistical data from Eurostat, OECD and national statistical offices.

    LinkedIn

    Frequently Asked Questions

    Further reading

    Explore our AI services

    Related reading

    Sources

    1. Eurostat — AI use in enterprises 2025 (DDN-20251211-2), December 2025(accessed 2026-04-15)
    2. Eurostat — ICT Enterprise Survey (ISOC_EB_AI), interactive databrowser(accessed 2026-04-15)
    3. OECD — AI use by firms continues to expand (announcement, January 2026)(accessed 2026-04-15)
    4. Stanford HAI — AI Index Report 2025(accessed 2026-04-15)
    5. UK DSIT / GOV.UK — AI Adoption Research (fieldwork Feb–May 2025)(accessed 2026-04-15)
    6. US Census Bureau — Business Trends and Outlook Survey (BTOS), Nov 2025 wave(accessed 2026-04-15)

    Next scheduled review:

    Ready to accelerate your AI journey?

    Book a free 30-minute consultation with our AI strategists.

    Book Consultation
    Share

    Get in Touch!

    The lab usually responds within 24 hours.

    Need help with AI?Get in touch