Natural Habitat.
AI Adoption in Numbers: Businesses, Trends and Studies
Statistics indicate growing use of artificial intelligence, but studies differ in the populations, periods and definitions of adoption they measure.
Eurostat data refer to European enterprises of a defined size. Results from international surveys should not be treated as directly equivalent.
Businesses in the European Union
According to Eurostat, in 2025, 20.0% of EU enterprises with at least 10 employees used AI technologies, up from 13.5% in 2024. That is an increase of 6.5 percentage points. For Poland the corresponding figures were 8.4% and 5.9%, a rise of 2.5 percentage points.
For AI adoption in enterprises, these numbers describe the spread of the technology in a defined population, not the proportion of successful projects.
What the indicator actually means
Eurostat measures whether an enterprise uses at least one AI technology. A business using AI for one task and an organization using it across several departments count equally as adopters. The indicator does not measure process maturity or savings.
When analysing AI applications, retain the year, geographical coverage and minimum company size. The findings cannot be applied directly to sole proprietorships.
Different samples, different results
The Stanford AI Index 2025 cites survey findings that 78% of respondents reported AI use in their organizations in 2024, compared with 55% in 2023. This differs from Eurostat’s measure of all enterprises in its statistical population. The research methods and populations are different.
When consulting on AI applications, comparing similar populations and research questions is more informative than searching for one universal adoption rate for the entire market.
Interpreting changes in the data
A change in percentage points is the absolute difference between two shares. A percentage change compares that difference with the initial value. Confusing the two can substantially alter the interpretation of the same trend.
Increased adoption does not by itself demonstrate better work quality or a profitable implementation. Assessing those outcomes requires process-level data on time, cost, errors and results.
Share of enterprises using AI
| Region | 2021 | 2023 | 2024 | 2025 |
|---|---|---|---|---|
| European Union | 7.7% | 8.1% | 13.5% | 20.0% |
| Poland | — | — | 5.9% | 8.4% |
Sources
- Eurostat, “20% of EU enterprises use AI technologies”, 11 December 2025, dataset isoc_eb_ai.
- Stanford Institute for Human-Centered Artificial Intelligence, “The 2025 AI Index Report”, 2025, Economy chapter.