Romania ranks last in the EU for AI adoption with its 5.2% share. This is the lowest in the EU, well below the EU average of 20.0% and far behind leaders like Denmark (42.0%), Finland (37.8%), and Sweden (35.0%).
The growth trend tells the real story. From 2024 to 2025, Romania's AI adoption rate rose by +2.1 percentage points. The EU average rose by +6.5 points in the same period. Denmark alone gained +14.5 points. Romania's growth is real, but it's not fast enough. While other countries are accelerating, Romania is falling further behind, not standing still.
The disparity becomes sharper when looking at company size. Across the EU, AI adoption rates are higher among larger companies: 55% of large enterprises use AI, compared to roughly 20% of medium-sized and 11% of small firms.
Romania follows the same pattern, but with lower figures across the board. The few Romanian businesses that have embraced AI tend to be large multinationals or ICT companies. Medium-sized enterprises, which are often the backbone of the economy, lag behind.
This is a "missing middle" problem. Medium-sized firms have the resources, but they're not the ones pushing AI adoption. They're stuck. And as a result, Romania isn't just lagging behind the EU average, but it's also falling behind its own potential.
If we zoom into specific sectors, it's the same story. AI is barely present in key parts of the economy. The information and communication technology (ICT) sector leads, followed by some banks and fintech startups, but for a lot of key industries, AI remains a distant concept. For instance:
- Manufacturing, which accounts for over 20% of Romania's GDP, has seen little beyond isolated pilot projects.
- Agriculture, where AI could drive huge improvements through things like precision farming, is still operating largely on traditional models.
- Healthcare has a few AI-driven initiatives in telemedicine and diagnostics, but these are limited in scale.
The digital foundation for AI is also weak. Only 11% of Romanian companies use cloud services, compared to the EU average of 34%. Big data analytics has just a 5% adoption, versus 14% across the EU. When you're not even using cloud or data analytics, it's hard to make a case for AI.
Skills shortages add another layer to the problem. Only 2.8% of Romanian employees work in ICT roles (half the EU average), and just 12% of companies provide any form of tech training to their staff (vs 22% in the EU). So even when there's interest in AI, there aren't enough people who know what to do with it.
The gap is not just technical. There is also a lack of trust. A lot of Romanian companies still see AI as a buzzword, something that sounds good on stage at a conference but feels risky or unnecessary in the day-to-day running of their business. Add to that the slow pace of policy execution, limited funding, and fragmented governance, and you have a situation where AI feels like a luxury most companies can't afford to think about.
On the policy side, the Romanian government adopted a National AI Strategy for 2024-2027 in July 2024, and the EU AI Act becomes applicable in August 2026. But Romania missed the August 2025 deadline to designate national competent authorities for enforcement, a sign that institutional readiness trails behind the paperwork.
Romania's AI gap is not the result of disinterest. It is a product of low digital maturity, uneven sectoral development, limited access to skills and infrastructure, and a persistent lack of strategic frameworks that make AI adoption viable for the majority of businesses.
