Europe has set three 2030 targets for business and skills: 75% of enterprises using cloud, AI or data analytics, more than 90% of SMEs at a basic digital level, and 20 million ICT specialists. Our analysis of Eurostat data shows that large organisations are already past the first line: 96.4% of Dutch enterprises with 250 or more staff qualify, and 98.1% of Belgian ones. For a large organisation the targets are not the goal. They describe the floor your peers already stand on, and the one item still moving fast is AI.
This piece is for the IT manager, CIO or informatiemanager at a Dutch or Flemish organisation of 250 to 5,000 people who is writing the 2027 roadmap now. Crux Digits builds AI systems alongside your ICT partner or internal IT team; we do not run infrastructure, and we hand over code you own.
What do Europe's 2030 digital targets ask of businesses?
Three things: 75% of enterprises using cloud, AI or data analytics, more than 90% of SMEs at basic digital intensity, and 20 million ICT specialists.
The targets come from Decision (EU) 2022/2481, the Digital Decade Policy Programme 2030. The Commission's DESI 2026 methodological note, dated 17 June 2026, defines how each is measured. The 75% counts enterprises that use any AI technology, buy sophisticated or intermediate cloud services, or perform data analytics. Basic digital intensity means using at least 4 of 12 listed technologies. The 20 million ICT specialists equal about 10% of total employment.
Progress is reported every June in the State of the Digital Decade report; the 2026 edition appeared on 17 June 2026. On the 2025 Eurostat survey, the EU as a whole stands at 63.2% on the combined measure, 46.7% on cloud and 20.0% on AI. Those are averages over every enterprise with 10 or more staff, and they hide the most useful fact for a large organisation.
Is a large organisation already past the 2030 line?
Yes, in almost every EU country: in 26 of 27 member states, at least 75% of enterprises with 250 or more staff already use cloud, AI or data analytics.
In 2025 the share was 96.4% for large Dutch enterprises, 98.1% for large Belgian ones and 93.4% for large German ones. Only Bulgaria, at 73.1%, sat below the line.
Cloud tells the same story. In Eurostat's cloud table, 88.5% of large Dutch enterprises and 91.1% of large Belgian ones bought sophisticated or intermediate cloud services in 2025, such as hosted ERP, CRM, security software or database hosting. In 19 of 27 countries, large firms are above 75%.
That is good news for a roadmap. The platform layer that most AI work depends on, hosted business systems with data you can reach, is already in place in most large organisations. The 2027 budget does not have to start with infrastructure.
Where is the gap that still matters in 2027?
AI in real processes: 67.6% of large Dutch enterprises and 76.4% of large Belgian ones used at least one AI technology in 2025, against 55.0% across the EU.
AI is also the fastest mover of the three. In Eurostat's AI table, the Dutch share among large firms rose from 41.7% in 2023 to 67.6% in 2025, and the Belgian share from 47.9% to 76.4%. Across all enterprises with 10 or more staff, the Netherlands added 9.6 points a year and Belgium 10.4. If that pace holds, our study finds large firms in 15 member states, the Netherlands, Belgium and Germany among them, at or past 75% AI use by the 2027 reference year.
Read that the right way round. In 2025, 32.4% of large Dutch organisations did not use AI. If yours is one of them, the peers who moved first now have systems you can learn from, and you can skip their experiments. If yours already uses AI, the 2027 question is different: how many of those systems run in a real process with an owner, a log and a budget line, rather than as a pilot or a licence.
Can you hire the AI skills a 2027 roadmap needs?
Not at the pace the targets assume: the EU had 10.45 million ICT specialists in 2025, and the 2023 to 2025 trend reaches 20 million only in 2052.
The Netherlands employs 708,800 ICT specialists, 7.2% of total employment and fourth in the EU, a number growing 2.6% a year according to Eurostat. Belgium is at 5.9%, growing 5.3% a year. Reaching 20 million across the EU by 2030 would take 13.9% growth a year; the fastest country, Portugal, managed 11.2%. Every large organisation in the Benelux recruits from that same pool.
The practical consequence is that AI work in 2027 has to fit the people you already have. Your ICT partner keeps running the estate. Your IT team keeps owning identity, data access and change. A specialist builds the AI system and hands it over in a form your team can support. That is a sourcing decision, and it belongs on the roadmap as one.
What should go on a 2027 IT roadmap?
Four moves, each tied to a finding: confirm the platform, take one AI process into production, source the build for supportability, and put an AI number in the board pack.
1. Confirm the platform, do not rebuild it. Finding: 96.4% of large Dutch enterprises already meet the combined target. Spend the first quarter listing which hosted systems you run (ERP, CRM, document management, the data warehouse) and which data in them an AI system may read under your existing access rules. The output is a short list of sources with an owner each, not a new platform.
2. Take one AI process from pilot to production by the third quarter. Finding: AI is the only one of the three technologies still moving fast, and large Belgian firms are already past 75%. Choose a process with a document flow and a measurable step, such as intake of invoices, contracts or service requests. Define what production means before you start: a named owner, logging, a fallback to a person and a line in the IT budget.
3. Source the build so your team can support it. Finding: Dutch ICT specialist employment grows 2.6% a year. Write into the contract that code and configuration are handed over, that your ICT partner joins the design review, and that your service desk receives a runbook. That keeps the AI system inside the change process and audit trail you already run.
4. Put a 2027 number in the board pack. Finding: a straight path to 75% AI use needs 49.9% of all Dutch enterprises in 2027 and 50.7% of Belgian ones; their recent pace gives 52.3% and 55.3%. Track your own organisation on the same Eurostat question, whether it uses at least one AI technology, and add the count of AI systems in production. The first number tells the board where you stand against peers. The second tells it whether the spend is working.
What does 2027 look like if the trend holds?
On a straight line through the 2023 and 2025 surveys, the EU crosses 75% on cloud, AI or data analytics in 2027, and the SME line crosses 90% the same year.
Our study labels these lines as arithmetic, not forecasts. The Commission's own baseline slows as adoption spreads, and puts the combined target in 2033 and AI for all enterprises not before 2035. Both readings point the same way: by the 2027 survey, AI in a large Benelux organisation is the norm, and the board's question moves on.
That shift is the positive part. The question stops being whether to use AI and becomes which processes run on it, who owns each system, and whether it passes the next audit. An IT leader who can answer those three in 2027 has a roadmap the board can fund, measure and defend.
Where does Crux Digits fit?
Crux Digits designs and builds AI systems alongside your ICT partner or internal IT team. We are not an ICT partner: we do not host, patch or run your estate. We scope one process, build the AI system and the software around it, document it for your change and audit process, and hand over code you own. The country detail behind this piece is in our analysis of Eurostat data on the 2030 targets and in our studies of AI adoption in the Netherlands and in Belgium.
If your 2027 question is how to take one AI system from pilot into a real process, see AI implementation in the Netherlands.
Last updated: 28 October 2026.
A consultant tells you where AI pays off; Crux Digits also builds it. A fixed price per step, one named expert, from Utrecht.
AI consultancy in the Netherlands →
