25 May 2026·6 min read·By Marcus Thorne

What EU business AI adoption means for you

EU business AI adoption hit 20% but the real story is the gap between Denmark's 42% and Romania's 5%. Here's why it matters for your business.

EU business AI adoption just jumped six and a half percentage points in a single year. Eurostat published the numbers last December. Twenty per cent of EU enterprises with at least ten employees now use AI somewhere in their business. Up from 13.5 per cent the year before. That sounds like momentum. And it is. But the headline buries a much stranger story.

The headline hides a brutal split

Denmark sits at 42 per cent enterprise AI adoption. Romania sits at 5.2 per cent. That is not a gap. That is two entirely different economies sharing a regulatory framework and a currency zone. The aggregate twenty per cent number flatters and obscures in equal measure. It averages out an economy that, on this question, no longer behaves like a single market at all.

If you are a developer in Copenhagen, you are working in an environment that outpaces the US enterprise average reported by Stanford. If you are in Bucharest, AI is still something that happens to other people. The adoption curve is not a curve. It is a cliff.

Capital is not on your side

Here is the deal. Roughly three-quarters of all AI venture capital in 2025 went to firms in the United States. The total: around $194 billion. The European Union, taken together, attracted $15.8 billion. Christine Lagarde cited these OECD figures in a speech to the European Parliament in November. They are not abstract. They are the reason a French SME looking at an AI pilot reaches first for a budget that does not exist.

And then reaches for a service that does. The service is almost always American.

Three clouds own the continent

Three US providers held roughly seventy per cent of the European cloud infrastructure market in 2025. European providers held about fifteen per cent. Every enterprise AI rollout that does not deliberately design around this reality ends up training on US compute. Billed in dollars. Governed by a foreign court's reading of data protection.

This is not hypothetical anxiety. The Next Web has documented it repeatedly.

Mistral's chief executive Arthur Mensch has spent the past year arguing that Europe must "own and operate" its own AI infrastructure.

Mistral put $830 million of debt behind a Paris data centre to make the point. A data centre is a long way from a continent-wide cloud. The structural dependency remains.

Your real bottleneck is people, not policy

The standard explanation for Europe trailing the US on AI is regulatory. The AI Act spooked boards. Tied up legal departments. There is something to that. But the deeper story is older and duller. Skills. Capital. Fragmented markets.

An OECD report from December 2025, prepared for the G7 presidency, asked SMEs what actually blocks adoption. Their answers are not the answers of executives frightened by Brussels. They are the answers of executives who would adopt AI tomorrow if they could find someone to install it, run it, and explain it.

The breakdown:

Fifty per cent cite a skills shortage as their primary barrier
Forty per cent point to maintenance costs
Thirty-two per cent flag hardware
Twenty-six per cent say they cannot understand the digital regulations they are meant to comply with

Real talk. If half your market cannot find talent and forty per cent cannot afford to keep the lights on, you do not have an enthusiasm problem. You have an infrastructure problem dressed in a policy debate.

Large firms live on another planet

Large enterprises in the EU adopt AI at around fifty-five per cent. Small ones sit at seventeen per cent. The gap is not philosophical. It is the difference between having a data engineer on payroll and not having one. Mario Draghi's earlier finding lands here: roughly seventy per cent of the per-capita GDP gap between the EU and the US is a productivity gap. Technology explains about two-thirds of that gap since the turn of the century.

These numbers are connected. Low AI adoption in small firms is a productivity story. A low-productivity economy cannot summon AI demand out of thin air.

The AI Act is late, not lethal

But that framing misses something. The Act's most invasive provisions, covering high-risk systems, do not apply until August 2026. The European Commission already moved to soften the edges. Its Digital Omnibus proposal from 19 November 2025 targets a twenty-five per cent compliance burden reduction overall, and thirty-five per cent for SMEs by 2029. It extended the simplified SME framework to firms with up to 750 employees and €150 million in turnover.

The Commission clearly read the survey data. Whether it read it in time is a different question. Industry analyses suggest EU and UK developers report launch delays in nearly six out of ten cases because of the Act. Roughly two-thirds of European companies still cannot articulate what their obligations are.

Regulation is not the main thing slowing EU business AI adoption. But it is not nothing. Pretending otherwise is dishonest.

Bright spots worth watching

Set against the unevenness, the bright spots are real and underreported. McKinsey's State of AI 2025 survey, with nearly two thousand respondents across 105 countries, found that 88 per cent of organisations globally now use AI in at least one function. Only six per cent see material enterprise-wide impact, defined as a five per cent or greater contribution to EBIT.

On that second measure, the European laggard problem is less severe than the headline numbers suggest. Americans are running more pilots. They are not necessarily running better ones. What separates high performers everywhere is not country but commitment: senior-leadership ownership, end-to-end workflow redesign, and a willingness to spend on infrastructure before measuring returns. Those are habits, not regulations. Europe can choose them at any time.

And some already have:

Siemens spent two years pushing its Industrial Copilot into factory-floor workflows, with new agentic capabilities announced at Automate 2025
SAP wove Joule into its core ERP
Mistral signed multi-year deployment deals with Accenture and at least one major European bank

What this means for you

The picture is not paralysis. It is unevenness with a shape. The firms doing AI well in Europe are large, well-capitalised, internationally minded, and concentrated in a handful of countries. The firms not doing AI are small, regionally bound, and disproportionately in the East and South.

If you work in tech, this split defines your opportunity map. The single market, on this technology, is two markets. Denmark and Romania are not on the same adoption curve. If the gap narrows, it will be because Europe finally treats AI adoption as a question of industrial policy and human capital rather than ethics frameworks. If it widens, the explanation is already sitting in the same survey data it has been sitting in for a decade.