The share of firms using AI in EU strategic sectors grew from 7% in 2021 to 18% in 2025, according to a new JRC analysis. While each of the 10 sectors advanced in use of AI, the growth was uneven, driven mainly by accessible, general-purpose AI — systems that analyse or generate text, code, images, sound and video.
The analysis and the methodology developed allow to measure and monitor the levels and development of AI usage, types of deployments and barriers to its adoption in the strategic sectors covered by the Apply AI Strategy.
Launched a year ago, the strategy aims to boost AI uptake and encourages organisations to consider the opportunities and risks posed by AI solutions to improve processes, efficiency and competitiveness. It proposes specific actions to support the adoption of relevant AI use cases in strategic sectors. The 10 strategic sectors are essential for EU economic security, green and digital transitions, and defence readiness.
They include agri-food, climate (climate and environment); defence (defence, security and space); electronic (electronic communications); energy, manufacturing (manufacturing, engineering and construction); media (cultural and creative sectors, and media); mobility (mobility, transport and automotive); pharma (healthcare, including pharmaceuticals) and tourism.
Considering the whole EU economic areas besides the strategic ones, the share of firms adopting the use of AI is slightly higher: it grew from 8% in 2021 to 20% in 2025.
The analysis drew on the Eurostat’s ICT enterprise survey data collected between 2021 and 2025, based on the EU classification of economic activities (NACE). It assesses adoption levels, technologies, applications and adoption barriers across them. The results are published in the report AI adoption across European strategic sectors.
What are the AI adoption trends in strategic sectors?
Firms in the cultural and creative sectors, media, energy, and electronic communications are leading the AI adoption with respectively 38%, 34% and 30 % of firms using AI. In these sectors, AI is used for a variety of business functions, from ICT security to R&D and production, demonstrating a broader diffusion of the technology.
In contrast, mobility, tourism and agri‑food are the sectors presenting the lowest uptake. In these, AI applications are still confined to marketing, sales, and other accessible tasks. Manufacturing and climate industries show a mixed pattern, combining ready-to-use AI tools with a modest shift towards using AI in production processes.
In general, the increased AI adoption is mainly driven by AI systems analysing and generating written or audio output for marketing and sales purposes, technologies which have become more accessible and are available at a lower cost.
From tools to core processes, different adoption patterns across strategic sectors
While the overall adoption curve is positive, the analysis highlights that in many cases AI still needs to be embedded into core business processes, such as production, where the greatest productivity gains are possible. Firms that tailor AI and integrate it into their production, workflows or infrastructure tend to achieve higher value-added outcomes, leveraging data and expertise.
The differences between capital-intensive industries, typically requiring sophisticated computing infrastructure, and labour-intensive sectors, based on people and skills, play also a role for understanding AI adoption. The data show that the capital-intensive sectors, such as energy and electronics, are experiencing a stronger growth in AI uptake than sectors more balanced in terms of capital and labour. Between 2021 and 2025, AI-uptake of energy firms grew from 18% to 34%, while electronics from 12% to 30%.
A sector-specific approach is needed to support AI adoption
The analysis highlights three key components when looking at how firms adopt AI: how much a company has reorganised its way of working to take full advantage of AI, how much it uses AI in production, and to what extent it is adopting ready-to-use AI solutions vs custom made ones, the level of AI commoditisation.
The ranking of strategic sectors along these key components shows a positive correlation between organisation and production. The firms investing more in structural organisational changes to integrate AI are more likely to use AI in production, suggesting that organisational changes must be prioritised. This is the case of energy, electronic communications or defence industries.
The analysis reveals that the use of ready-to-use AI solutions vs custom-made ones depends on the specificities of each sector. For instance, the climate sector, despite having a medium AI uptake, displays a low uptake of off-the-shelf AI solutions, as it requires a higher degree of customisation compared to other strategic sectors.
At the same time, the manufacturing sector, overall displaying a low AI uptake, ranks higher than climate for adopting off-the-shelf solutions. This indicates that using accessible ready-to-use AI solutions does not necessarily increase integration of AI into production processes.
Overall, the analysis suggests that policy measures aimed at increasing AI integration into core processes need to be tailored to the specificities of the strategic sector.
Bottlenecks to AI adoption
The data show that the lack of relevant expertise is widely indicated as the main obstacle to AI adoption across the vast majority of strategic sectors. In particular, two factors emerge as the dominant bottlenecks: the scarcity of specialised skills and the regulatory uncertainty, including concerns about data protection and privacy. The report suggests that digitalisation initiatives, skills development, and organisational change can support AI uptake and close the gap between tool uptake to deeper integration.
The results of the study will support policy makers with setting targets and create new sector-specific measures on how to increase AI adoption in the strategic economic areas.
Related Links
AI adoption across European strategic sectors
Policy Brief AI adoption in the European Union – New evidence for Apply AI strategic sectors
