I just noticed that even here I can directly generate a post using AI. I don’t have to copy and paste. Now I am not sure if blogging still makes sense nowadays. Nevertheless, I will do something else. I will write something that AI is not going to generate, something that is mine. Actually, I used AI to prepare some information and to gather some data from documents. But the questions I asked are 100% mine.
When I was sitting in Copenhagen in the Cafe 2 weeks ago, one question came into my mind: What is AI adoption in this country in comparison to Poland? The answer that Denmark has significantly better adoption didn’t surprise me a lot. More surprising was the data about Poland. In that moment, I realised I live in a bubble. I thought that we Poles use AI more and we are more eager to adopt AI in companies. Thankfully, I am no longer stupid and ignorant about this concrete topic. This thinking process forced me to ask more questions, taking into account statistics and reports. In this post, I will just share what I discovered.
One digression: I am a huge fan of AI, but I hate AI-generated images. Nevertheless, this time I used AI to combine my photos from Copenhagen to create an image for this post.
Here are my questions and answers that I found out while digging deeper.
1. Why does Denmark have such high AI adoption among companies and individuals?
My first thought was “money”. Denmark is just rich, so it is easier to build AI solutions that are not cheap. But how to explain why Estonia develops technologies faster and more eagerly than Germany, while Germany is the country that has more financial resources than Estonia? So it was a bad assumption. Money helps, but it’s not the reason.
Denmark has built several complementary advantages.
- Digital skills – around 81% of Danes had at least basic digital skills – the EU’s average is 60%
- A lot of companies, even before the AI wave, were digital
- Labour is expensive – I have to admit I ignored this aspect at first, but it makes sense. When employees’ salaries are high, the products that the company produces are expensive, which makes the company less competitive. Companies have to cut costs somehow to take part in the race. The best way is to cut the most expensive costs.
- Public digitalisation normalised technological change – Denmark spent a lot of time before AI to make interaction with government digital.
Danes, nevertheless, are not super optimistic about the technology. They clearly say (over 80%) that AI should be regulated and should be considered a priority.
2. What makes Europeans in some countries more eager to adopt new technologies?
If you have for around 15 years digital banking, digital identity, online tax administration, electronic prescriptions, cashless payments, fast internet, and digital government, well-designed apps, and you have in mind “new digital technology usually makes my life easier,” this creates technological confidence. People used to technologies that really increase the level of living aren’t afraid so much about new technologies.
3. What is the main difference between Germany and Poland in AI adoption?
Germany is much better at organisational adoption. When we take into account the size of the company, there is no significant difference when we compare large companies. When we compare SMEs, the gap between Germany and Poland is huge (which personally made me sad, but I also see the business opportunities).
Why is there such a big difference? I asked ChatGPT to explain. Here is the answer: “Germany’s AI story is much more about AI entering existing business and industrial systems than creating huge consumer excitement. Poland has strong technical individuals but weaker organisational diffusion”
4. Why don’t Polish SMEs adopt AI faster?
The answer is super simple. But at upcoming conferences, I am going to ask business owners on my own. So the answer is “They don’t understand AI”. They probably don’t see the opportunities at all, so I hope I will be able to help someone understand it better to save some money and time 🙂
There are some smaller things as well. Small companies often don’t have a structure: documents in different formats, pricing in Excel, inconsistent data, no APIs, ERP with outdated information. Next thing is that large companies have people dedicated to adopting AI; small companies don’t, which makes things more difficult to deploy.
5. What are the most popular AI use cases among European SMEs?
The most common applications are much more boring than autonomous AI agents running entire companies.
They include:
- text analysis,
- writing and content generation,
- marketing,
- sales,
- customer service,
- administration,
- document processing,
- internal knowledge search,
- programming,
- data analysis,
- forecasting,
- accounting support.
Example: Germany
Among German enterprises using AI, the most common technologies in 2025 included:
- text mining: 52%,
- image/video/audio generation: 52%,
- speech recognition: 42%,
- natural-language generation including code: 35%,
- workflow/decision automation: 27%,
- machine learning: 23%.
Example: Finland
Finland shows a similar structure.
Among all Finnish enterprises in 2025:
- 29% used AI to analyse written text,
- 22% generated images/video/audio,
- 20% generated text, speech or code.
More interesting non-IT examples
Outside IT, examples include:
- a technical wholesaler automatically preparing quotations from incoming emails,
- an industrial service company searching manuals and previous repair cases,
- a manufacturer detecting defects through computer vision,
- a logistics company predicting volumes and optimising routes,
- an accounting firm classifying invoices,
- a law firm analysing contracts,
- a hotel automatically answering recurring customer questions,
- a real-estate company qualifying leads,
- a machinery company predicting maintenance needs.
This is where I think the next adoption wave becomes interesting.
The first AI wave was: ChatGPT for employees.
The second is: AI embedded inside workflows.
6. Do Europeans feel AI gives them more free time or higher productivity?
The strongest narrative is productivity, efficiency, and growth, not leisure.
The answer to this question is usually framed as freeing up time for higher-value work. In the case of programmers, it’s using AI to do repetitive, low-value work to free employees for the tasks where AI can’t really help.
To summarise, the leftist dream about a 4-day working week is not going to happen thanks to AI, the goal is to do less repetitive work, make faster decisions, and gain higher output per employee.
7. Which industries lead AI adoption in Europe?
Including IT
The winner is overwhelming: Information and communication
EU: approximately 63%
Finland: 80%
Sweden has recently reported extraordinarily high adoption among IT/communication businesses as well.
This includes:
- software companies,
- telecom,
- cloud businesses,
- digital platforms,
- data companies.
It is not surprising.
AI is close to their core competency.
Excluding IT
1. Professional, scientific and technical services
EU adoption is around 40%.
This includes:
- consultancies,
- engineering,
- legal/accounting activities,
- R&D,
- technical services,
- specialist professional companies.
Why? Because much of their product is essentially: information + expertise + documents + decisions. LLMs are particularly good at assisting exactly these activities.
2. Manufacturing
EU average is only around 18%, but advanced manufacturers are interesting because applications can have enormous value:
- predictive maintenance,
- computer vision,
- quality control,
- production optimisation,
- demand forecasting,
- engineering assistance,
- digital twins.
3. Finance
Finance is also naturally AI-friendly:
- fraud detection,
- risk modelling,
- AML,
- document analysis,
- customer support,
- credit decisions,
- algorithmic trading,
- software development.
4. Pharma / biotechnology
AI is particularly strong in:
- drug discovery,
- molecular modelling,
- clinical data,
- protein prediction,
- trial optimisation.
To be honest, number 4 is the most abstract for me. I am not even able to imagine properly how AI works in this industry, but it sounds promising.
8. Why does Norway perform so well? Is it the same case as Denmark?
Mostly, yes, but Norway has an interesting additional feature: the state itself is a powerful technological adopter.
Norway combines:
- high incomes,
- very high labour costs,
- high institutional trust,
- a small population,
- strong digital skills,
- high-quality government data,
- excellent digital infrastructure,
- advanced public services.
Its national AI strategy explicitly treats high-quality public data and trust as national competitive advantages. Norway also has enormous state financial capacity because of its sovereign wealth. But oil wealth alone cannot explain AI adoption.
You cannot just easily purchase:
- institutional trust,
- digital skills,
- high-quality historical datasets,
- competent administration.
These were built over decades.
9. Which industries have the lowest AI adoption in Europe?
Transport and storage
EU: approximately 11%
Finland: 13%.
Construction
EU: approximately 11%
Finland: 24%, which is actually quite high relative to the EU average.
Other relatively weak areas include:
- hospitality,
- parts of traditional retail,
- very small manufacturing businesses,
- personal services.
10. Can we already see significant GDP growth caused by AI?
Not yet. At least we cannot convincingly isolate it.
“The IMF’s 2025 European modelling estimates that AI could raise European productivity by approximately 1% cumulatively over five years, although results vary strongly by country and scenario.
The OECD estimates possible additional annual total-factor-productivity growth of roughly 0.25-0.6 percentage points over a longer horizon under its scenarios.”
11. Which European countries lead in successful AI startups?
In biggest scale:
- United Kindom
- France
- Germany
But taking into account smaller but very strong ecosystems
- Sweden
- Switzerland
- Netherlands
- Finland
- Denmark
- Norway
- Estonia
12. Why are those countries successful?
- Excellent technical talent
- Capital
- History of successful companies
- Government support – it’s not only about money, but it’s also a social system as well that encourages entrepreneurs to take risks. If something goes wrong the government will support to not make you homeless like it happens in some other countries in Europe when you want to risk a little more than others.
- Ambition to go global
- Access to sophisticated customers
13. What does Europe’s lifelong-learning culture look like statistically?
Lifelong learning culture is one of the factors that makes people more open to learning new things. That’s why some countries do better with AI adoption. I heard about this term for the first time while digging deeper into AI. This terms says in my opinion a lot. If people don’t have the constant learning habit, they do not participate in online and on-site training often, they are cursed with failure. They may not even be aware of what kind of opportunity is passing by.
High lifelong-learning culture:
- Sweden
- Denmark
- Finland
- Netherlands
- Estonia
Middle:
- France
- Austra
- Belgium
- Spain
- Germany
Weak:
- Poland
- Italy
- Greece
- Romania
- Bulgaria
- Croatia
14. How do successful countries build trust in technology?
- Technology must actually work
- Citizens need to feel protected
- People need digital competence
- Institutions need to admit risks
- Make benefits visible – digitisation has to produce tangible benefits
- Involve citizens rather than impose everything from above
Summary
To make AI adoption possible, we have to take many factors into account. Unfortunately, all factors have to perform relatively well. If at least one is closer to zero, the adoption of new technology is not going to happen smoothly or will be really difficult. Success is like a product of the multiplication of all factors. If one goes to zero, the entire equation goes to zero.
Many people are now afraid of loosing job because of AI, I see the opportunity for many years and many people. The key point is that we have to be open to learning new things, and we have to build more trust that technology is going to help us. Change is a normal part of our lives. Sole constant thing in life is change.
Read my previous posts as well.
Sources
Eurostat and European Commission
These are the most important ones.
- Eurostat: Digitalisation in Europe 2026, probably the best single starting point. It covers AI adoption by businesses, company size, digital skills and individual GenAI usage. In 2025, 20% of EU enterprises used AI, versus 55% of large enterprises and 19% of SMEs.
- Eurostat: Enterprises using AI technologies, 2025, the underlying country comparison, including Denmark 42%, Finland 38%, Sweden 35%, Germany 26%, Poland 8.4% and Romania 5.2%.
- Eurostat: 20% of EU enterprises use AI technologies, a shorter summary of the 2025 enterprise data.
- Eurostat: Key Figures on European Business 2026, particularly useful for comparing SMEs with large companies.
- Eurostat: Digitalisation in Europe 2025, useful for comparing the 2024 and 2025 adoption waves.
- European Commission: Digital Decade 2026 country reports, where you can open the individual report for every EU country.
- Poland Digital Decade Report 2026, especially relevant because it explicitly identifies business digitalisation, SME technology adoption and digital skills as Polish weaknesses.
- Denmark Digital Decade Report 2026
- Finland Digital Decade Report 2026
- Sweden Digital Decade Report 2026
AI adoption outside Eurostat
- OECD: Empowering SMEs in the Age of AI, 2026, one of the best reports for our discussion about opportunities in SMEs. It covers more than 2,000 SMEs across 12 OECD countries and examines tools, barriers and motivations.
- EIB: How EU firms are faring with AI and big data, based on around 13,000 EU firms. Particularly interesting because it compares Europe with the US and looks at GenAI integration into actual business processes.
- Microsoft: Global AI Adoption in 2025, useful because its methodology is based partly on observed usage rather than simply asking people whether they use AI.
- Stanford AI Index 2026: Economy, much broader than Europe, covering investment, adoption, productivity, employment and AI companies.
- IMF: AI and Productivity in Europe, particularly useful for question 10 about whether AI is already affecting GDP and productivity.
Sweden and Finland
These are excellent because the national statistical offices provide considerably more detail than the headline Eurostat numbers.
- Statistics Sweden: AI in enterprises 2025
- Statistics Sweden: AI use in enterprises full report
- Statistics Sweden: AI in Sweden 2026 update, especially useful for the IT vs non-IT distinction. For example, AI usage reaches 89.6% in IT-related activities but only 12.3% in transport/storage.
- Statistics Finland: AI use by Finnish enterprises 2025, including sector-level data. AI adoption reaches 80% in information and communications.
Denmark and SMEs
- SMVdanmark: AI is not yet everyday reality for Danish SMEs, particularly interesting because even in Denmark, 74% of SMEs in this survey had not integrated AI.
- Danish Technological Institute: Knowledge gap slowing AI in manufacturing, where 65% of manufacturing SMEs are classified as AI-waiting and only 10% as AI frontrunners.
Startups
- Dealroom: European AI agent startups 2026, with companies such as Legora, ElevenLabs, Parloa, Lovable, n8n, H and others.
- Sifted AI 100 Europe 2025, a particularly good source for discovering European AI startups by industry rather than simply looking at the largest funding rounds.
Lifelong learning
- Eurostat: Adult Education Survey
- Eurostat: Adult learning across European regions
- Eurostat: Adult participation in education and training, the source behind the interesting result that Sweden reached 73.9% in the 12-month measure versus 46.6% across the EU and 24.3% in Poland.
Trust and Norway
- Norway: National Strategy for Artificial Intelligence, interesting because Norway explicitly identifies public trust, digital competence, infrastructure, registry data and cooperation between government, employers and unions as competitive advantages.
- Norway: Trustworthy AI strategy, specifically about maintaining trust through privacy, human rights, cybersecurity and responsible AI.
The company case studies
These are useful if you want examples rather than statistics.
- PostNord NIVA AI customer-service case, about 10,000 calls/day and 20 to 25% resolved by the AI system.
- KONE + Helsinki University Hospital predictive maintenance, AI analysing hundreds of elevator parameters and detecting anomalies before failures.
- KONE 24/7 Connected Services results, including 70% more proactive maintenance actions and 30% fewer end-user-visible call-outs in its reported first-two-year results.
- Wärtsilä Expert Insight: Aurora Spirit case, anomaly detection in a ship engine before a serious failure.
- Danfoss + Vantaa Energy AI heating optimisation, predictive control deployed across 300 buildings.
- Maersk AI-enabled warehouse robotics, including order sorting three times faster than conventional systems in the reported implementation.
- IKEA Finland Aava chatbot
- Academic study: AI development with 11 Finnish SMEs, especially interesting for understanding how SMEs actually progress from business problem → data → experiment → ML/AI solution.
If you wanted to read only five, I would choose Eurostat Digitalisation 2026, OECD SMEs & AI 2026, EIB Investment Survey, Statistics Sweden 2025/26 and the Finnish SME academic study.
