Where EU IT budgets are actually going in 2026, and why automation is the biggest line item.

European IT spending crossed into genuinely large numbers in 2026, and most of the coverage of that number stops at the headline. The more useful question is where it is actually going: how much is public money versus enterprise budgets, how much is infrastructure versus compliance, and how far real automation adoption actually sits behind the spending that is supposed to be funding it. The short version is that the money and the adoption curve are not moving at the same speed — and a regulatory deadline that just took effect changes what "automation" is allowed to mean for the rest of this year.

How far along the EU actually is on its own 2030 targets

The Digital Decade programme set four measurable targets for 2030. Each bar below is how much of its target has been reached as of the latest figures — the raw numbers behind each one are in the panel. Plotting them on a shared scale is what makes the real story visible: the technology targets are within reach, and the people ones are not.

The 2026 numbers, side by side

The two EU programme figures are commitments over a multi-year period, not single-year spend — they are not directly comparable to Gartner's annual enterprise total, but they show where the public half of this spending is aimed.

The headline number, and what is actually driving it

Gartner's own forecast puts European IT spending at $1.4 trillion in 2026, up 11.1% on 2025 — a growth rate well ahead of the broader economy, and one Gartner attributes specifically to AI, cloud and cybersecurity investment rather than a general IT refresh cycle. The clearest single line item behind that growth is infrastructure: end-user spending on AI-optimised servers is forecast at $46.8 billion in 2026, up from $39.3 billion the year before. That is not a rounding change — it is the physical hardware European organisations are buying specifically to run AI workloads, and it is growing faster than IT budgets overall.

The practical read for a business outside that infrastructure layer: the growth is real, but it is concentrated. A company is not seeing 11% more useful software land in its own stack this year — it is watching a smaller number of organisations buying GPU clusters and hyperscale capacity pull the average up, while everyone else's actual software and services spend grows closer to a normal year. Treat the headline figure as a market-heat indicator, not a benchmark for your own budget.

Where the public half of the money is aimed

The EU's own dedicated digital infrastructure fund, the Digital Europe Programme, runs €7.5 billion across 2021–2027, targeted at five areas: supercomputing, artificial intelligence, cybersecurity, advanced digital skills, and the wider deployment of digital technology across the economy and public administration — the last of which covers things like interoperable government services, not consumer-facing product. It is deliberately not a grants-for-startups fund; it is closer to shared infrastructure the private sector is expected to build on top of.

That is a fraction of what member states are actually committing on their own. Under their National Digital Decade Roadmaps, EU governments have pledged 1,934 measures worth €289.3 billion, of which €205.9 billion comes directly from public budgets — roughly 1.09% of EU GDP. The European Commission's own accounting puts real economic weight behind that spend: it estimates every euro invested returns €1.50 in EU output, and €2.00 globally, by the end of the Digital Decade. Whether that specific multiplier holds is a separate question — the scale of the commitment itself is not in doubt.

The deadline that just changed what "automation" has to account for

On 2 August 2026, the EU AI Act’s Article 50 transparency duties — chatbot disclosure, AI-generated content marking, deepfake labelling — and the Commission’s enforcement powers over general-purpose AI models took effect. That part of the timeline held. What did not hold is the harder deadline everyone had been planning around: the Council and Parliament’s "Digital Omnibus on AI," finalised weeks before, pushed the Annex III high-risk obligations — the rules covering AI used in hiring, credit scoring and similar consequential decisions — from 2 August 2026 to 2 December 2027, a sixteen-month deferral.

The detail worth building around is a narrower one: the same package clarified that AI features used purely for internal automation, performance optimisation or user convenience are no longer automatically classified as high-risk safety components. That distinction matters directly for the kind of automation most businesses are actually deploying — a workflow tool that routes support tickets or drafts a first-pass email reply is not the same regulatory category as a system deciding who gets a loan, and the rules now say so explicitly rather than leaving it to a case-by-case reading. It does not remove the need for basic AI governance, but it meaningfully narrows what needs the heaviest compliance machinery.

Adoption is real, and much smaller than the spending suggests

Eurostat's own count, published at the end of 2025, is specific: 20.0% of EU enterprises with 10 or more employees used AI technologies, up from 13.5% the year before — a genuine 6.5-point jump, not a rounding artefact. But that average hides a wide spread. By company size, it is 55.0% of large enterprises against 30.4% of medium and just 17.0% of small ones. By country, Denmark (42.0%), Finland (37.8%) and Sweden (35.0%) lead; Romania (5.2%), Poland (8.4%) and Bulgaria (8.5%) sit near the bottom of the same union.

The gap by sector is even starker: information and communication companies lead at 62.5% adoption, professional and technical services follow at 40.4%, and most other sectors sit well below the EU-wide average. If your own sector or country is not one of the leaders on that list, being behind "the EU average" is not actually being behind your real competitors — it is being roughly on pace with most of the union outside a handful of concentrated hubs.

The foundation is ahead of the automation, and that is the right order

The Commission's own 2026 Digital Decade figures show cloud computing at 46.7% enterprise adoption and data analytics at 39.9% — both more than double the 20% AI figure. That is not businesses falling behind on AI; it is the normal build order. Meaningful automation depends on having clean, accessible data and infrastructure to run against first, and the EU's adoption curve is following exactly that sequence: cloud and data foundations ahead, AI layered on top once they exist.

The practical implication cuts against a common instinct: a business tempted to buy an AI automation tool before its underlying data is in a workable state is skipping the step the EU-wide numbers show almost everyone else did first. A Technical Audit that actually checks whether the data an automation would depend on is clean and accessible is a cheaper mistake to catch than a stalled automation project six months in.

The constraint the money cannot buy

Spending is the easy half. The EU employed 10.4 million people in ICT specialist occupations in 2025 — around 5% of total employment — against a Digital Decade target of 20 million by 2030. That is not a shortfall to be closed by a growth spurt; it is a doubling in five years, and the Commission’s own 2026 State of the Digital Decade assessment warns the target will be missed without a change of pace. The talent pool is also narrow in a way that limits how fast it can grow: 19% of those specialists are women against 81% men.

The broader base is no easier. Around 56% of EU citizens aged 16 to 74 have at least basic digital skills, against an 80% target for 2030. For a business, that is the number with the most immediate operational meaning in this whole article: an automation project fails on adoption far more often than on technology, and the people expected to use the new system are drawn from that same distribution. Budget for training and change management as part of the automation, not as an optional extra after go-live — the spending figures show what is being bought, not who is available to run it.

Where SMEs actually sit on this curve

The headline spending is enterprise-weighted, so it is worth separating out where smaller businesses genuinely are. Eurostat puts 71% of EU SMEs at or above a basic level of digital intensity in 2025, against a Digital Decade target of more than 90% by 2030 — a real majority already over the line, but a 19-point gap with five years left. "Basic digital intensity" is a deliberately low bar: it counts things like having a website, using cloud services or e-invoicing, not AI or automation.

That framing is useful precisely because it is unglamorous. A business that has not yet cleared the basic bar gets more measurable return from doing so — a site that actually converts, invoicing and CRM that talk to each other — than from an AI layer on top of processes that are still manual underneath. If the basics are already in place, the same figures say the next step is not a general "AI transformation" but the one process costing the most hours, which is the shape a Workflow Automation engagement is built around.

How we scope this in practice

Most automation work that actually ships starts narrower than "AI Act compliant, enterprise-wide transformation" — a Workflow Automation engagement replacing one specific manual process (ticket routing, lead qualification, report generation) with a deterministic pipeline and an AI step only where judgment is genuinely needed, which is also the pattern least likely to land in the AI Act's higher-risk categories in the first place.

Where the goal is closer to what the EU-level spending is actually funding — genuine infrastructure and process change, not a single tool — that is scoped through AI & Automation as its own engagement, built around whichever systems your business already runs rather than a generic package.

Sources

The spending, adoption and AI Act figures above come from these, checked August 2026.

• Gartner — Gartner Forecasts IT Spending in Europe to Grow 11% in 2026 ↗

  Total European IT spending forecast at $1.4 trillion in 2026 (+11.1% YoY), attributed to AI, cloud and cybersecurity investment; AI-optimised server spend at $46.8 billion, up from $39.3 billion in 2025.

• European Commission — 2026 State of the Digital Decade package ↗

  EU enterprise adoption: 46.7% cloud computing, 39.9% data analytics, nearly 20% AI. National Digital Decade Roadmaps: 1,934 measures worth €289.3 billion, including €205.9 billion from public budgets (around 1.09% of EU GDP).

• European Commission — Digital Europe Programme: €7.5 billion of funding for 2021-2027 ↗

  The EU's dedicated digital infrastructure fund and its five priority areas: supercomputing, artificial intelligence, cybersecurity, advanced digital skills, and wide deployment of digital technology across the economy and public administration.

• Eurostat — 20% of EU enterprises use AI technologies ↗

  AI adoption among EU enterprises with 10+ employees rose from 13.5% in 2024 to 20.0% in 2025. Breakdown by size (17.0% small, 30.4% medium, 55.0% large), by country (highest: Denmark, Finland, Sweden; lowest: Romania, Poland, Bulgaria) and by sector (highest: information and communication, 62.5%).

• Eurostat — Towards Digital Decade targets for Europe ↗

  2025 figures against the 2030 Digital Decade targets: 10.4 million people employed in ICT specialist occupations (about 5% of EU employment) against a 20 million target; 19% of ICT specialists are women; 71% of EU SMEs at or above basic digital intensity against a target of more than 90%; and around 56% of people aged 16–74 with at least basic digital skills against an 80% target.

• Council of the European Union — Artificial Intelligence: Council gives final green light to simplify and streamline rules ↗

  The "Digital Omnibus on AI," finalised 29 June 2026, defers Annex III high-risk AI obligations from 2 August 2026 to 2 December 2027, while leaving the Article 50 transparency duties and GPAI enforcement powers taking effect on 2 August 2026 unchanged.