Making AI work for Europe
Ipsos has released a new report, commissioned by Google, examining the barriers that European businesses face in adopting AI and how those barriers could be overcome, in light of the EU’s policy objective to leverage AI for economic growth and competitiveness.
Key findings
The adoption landscape is uneven:
- Eurostat data indicates that while approximately 20% of EU businesses used AI technology in 2025, this masks significant variation by sector, size and region.
- AI use is heavily concentrated in digitally mature sectors: 62% of information and communication businesses use AI tools, compared to just 17% of manufacturers and around 11% of construction firms.
- Large businesses are considerably more likely to adopt AI than SMEs, with this gap widening from 30 percentage points in 2024 to 38 percentage points in 2025.
Human-centric barriers matter as much as, if not more than, technological ones:
Although AI has created some fear and uncertainty for many businesses and workers, our analysis suggests that many of the issues individuals are grappling with when thinking about AI are similar to those in past waves of technology. In that context, we identify six core questions that individuals must navigate before deciding to adopt AI:
- Is AI relevant to me?
- What is the value of AI?
- What are the risks?
- What are the rules?
- Can I trust AI?
- Do I have the capability to use AI?
Bridging the adoption gap requires a human-centric approach:
Our report outlines three guiding principles for policymakers, business leaders and technology providers
- Guide sectors to identify task-specific relevance to build the case for AI's utility and value through sector-specific dialogue, visible case studies and regulatory clarity.
- Support organisational capacity and readiness, particularly among SMEs, to enable experimentation and move beyond pilots to full-scale implementation.
- Build AI literacy at all levels to ensure organisations have the skills mix to capitalise on AI opportunities safely and effectively.
Technical note:
This report draws on a rapid evidence assessment of approximately 70 studies into AI adoption across Europe, and 15 semi-structured interviews with senior stakeholders engaged with AI adoption issues across EU Member States, including Austria, Belgium, Croatia, Denmark, Estonia, France, Italy, Latvia, Netherlands, Poland, Portugal, Spain and Sweden. Interviews were conducted between January and March 2026. All studies reviewed were published after 2024 to ensure relevance to recent trends in generative AI.
