Artificial intelligence (AI) use in healthcare is moving ahead rapidly. The WHO Regional Office for Europe (WHO Europe), has produced an informative report co-funded by the European Commission. EHTEL has examined the report in depth, and translated the survey results into five key findings. In it there are the implications for implementers both today and in the immediate future. EHTEL adds further value by identifying three top priorities for digital health leaders.

Source: WHO Europe

From experimentation to implementation

AI is no longer a future possibility for health systems. Instead, AI is becoming a core component of healthcare delivery, governance and innovation. Recognising this shift, WHO Europe has published the Artificial Intelligence is Reshaping Health Systems: State of Readiness Across the European Union  report. The report is based on a first, comprehensive 2024-2025 survey assessing AI readiness across the WHO European Region. The survey achieved an exceptionally high participation rate of 94% of the respondents approached. All 27 European Union (EU) Member States contributed with their priorities and experiences.

The report arrives at a pivotal moment. The EU has adopted the world’s first comprehensive AI legislation in 2024 (the AI Act) and is preparing for the act’s entry into force in late 2027. Simultaneously, the Union is getting ready for the March 2029 implementation of the European Health Data Space (EHDS). These two developments create an unprecedented policy environment for scaling trustworthy AI in health. A shift is definitely taking place in the move in healthcare of AI from experimentation to implementation.

For digital health implementers, the message is clear. Europe is entering the era of AI deployment, but readiness remains uneven and significant implementation gaps persist.

Source : EvalCommunity Academy

Five findings for the attention of every implementer

The report contains five key findings:

• Strategy is widespread, but health-specific planning is still rare
• Workforce readiness is the biggest hidden challenge
• Data readiness is improving, but interoperability remains a bottleneck
• AI adoption is already happening, and diagnostics are leading the way
• Trust and governance are becoming the decisive success factors

Strategy is widespread, but health-specific planning is still rare

The good news is that AI is firmly on the political agenda:

• 85% of EU countries (23 of 27) already have a national cross-sectoral AI strategy.
• 80% of countries identify health as a sector that will be significantly affected by AI.

However:

• Only 11% of countries have developed a dedicated health-specific AI strategy. See the box (below) for information on Finland, Slovakia, and Sweden.

• Another four countries (Cyprus, Czechia, Italy, and Spain) are developing a health-specific AI strategy.

Many healthcare organisations today are therefore (still) expected to implement AI solutions without a tailored national roadmap being in place that would address clinical workflows, patient safety, procurement, or workforce implications.

Workforce readiness is the biggest hidden challenge

Perhaps the most striking finding of the report is the limited investment in AI competencies on the part of countries:

• Only 26% of countries provide in-service AI training.
• Only 22% of countries offer pre-service AI education.
• Merely 15% of countries provide both pre-service and in-service AI education.

At the same time, fewer than half of EU countries (48%) have created dedicated professional roles in the fields of AI and data science in health systems. The result is a growing gap between technological ambition and organisational capability.

Data readiness is improving, but interoperability remains a bottleneck

AI depends on access to high-quality, interoperable data. The report shows encouraging European progress in this field and data readiness is improving:

• 67% of countries have a health data strategy.
• 63% of countries have established a national health data hub.

However:

• Only 44% of countries have a formal health data governance framework.
• Only 33% of countries have rules governing cross-border health data sharing.

The interoperability bottleneck is particularly important to handle now, because the EHDS is moving towards a more intensive implementation phase in 2029.

As additional background:

The work performed by the Joint Action TEHDAS2 – in the context of secondary use of data – has allowed countries to already make important progress, even if it is not yet complete. It does, however, include a first framework proposal for enhanced collaboration. This framework is essential since it now provides European Member States with a real foundational benchmark on which to build.

An EHTEL briefing paper  also makes another important point. AI is increasingly becoming a key enabler for the FAIRification of data. In itself, data FAIRification is also an important condition for the use of AI supported decision and monitoring systems.

AI adoption is already happening, and diagnostics are leading the way

The survey confirms that, in many European health systems, AI is no longer handled experimentally.

The current AI applications being used by countries chiefly include diagnostics and conversational AI/chatbots:

• 74% of countries are using AI-assisted diagnostics.
• 63% of countries are using conversational AI or chatbots for patient support.

Countries’ main motivations for the use of AI are also remarkably consistent across Europe:

• 100% of countries see AI as a way to improve patient care.
• 96% of countries view AI as a tool to reduce workforce pressures.
• 89% of countries expect to obtain gains in efficiency.

Yet there is an important caveat: it is financing. Of the countries that identified priority areas for AI, only 63% had allocated dedicated funding.

Trust and governance are becoming the decisive success factors

Positively:

• 63% of countries cite guidance on transparency, verifiability and explainability is the most important policy enabler.
• 56% of countries consider clear accountability and liability rules to be essential for wider adoption.

With regard to other decisive success factors, technology itself was not perceived by European countries as the main barrier to AI adoption.

Rather, the report identifies as financial affordability as the most significant implementation barrier.

Interestingly, no country considered environmental impact to be a major barrier. The report highlights, however, that environmental impact is an emerging issue that requires future attention.

While infrastructure and usability have often been highlighted – as in the EHTEL Briefing Paper supported by the Xt-EHR project – as urgent issues to be addressed by most advanced implementers, it is noteworthy that, in the WHO Europe report, no country identified these two as critical issues. The WHO Europe report did, however, remark in its conclusions that “[h]igh upfront costs – including technology, infrastructure and workforce training – pose a particular barrier for resource-constrained health systems.” (p42).

Three top priorities for digital health leaders

Expanding on the overall report results, three key priorities emerge for health system implementers. Several these ideas – e.g., around data quality, risk management, accountability, transparency, and usability – have their foundations in EHTEL’s 2026 Briefing Paper:

Build AI-ready organisations: Invest in workforce development, create multidisciplinary AI teams, and integrate AI competencies into existing digital transformation programmes.

Strengthen data foundations: Prepare for EHDS implementation by improving interoperability, data quality, and governance frameworks that enable secure data sharing and secondary use.

Embed trust by design: Implement robust governance processes which cover – from the earliest stages of deployment – risk management, evaluation, transparency, and accountability.

The road ahead for implementers

Europe is entering a new phase in digital health. The policy foundations are being laid through the AI Act and the EHDS and their upcoming implementation. Many countries are already deploying AI applications at scale.

The WHO Europe report demonstrates, however, that readiness remains uneven – particularly in areas such as workforce capacity, governance, and implementation financing.

For digital health implementers, the central lesson is simple. The fundamental question is no longer whether AI will become part of health systems. Instead, we need to ask ourselves whether health systems are building the capabilities needed to implement AI responsibly, equitably and sustainably.

To conclude, it is the countries that invest now in people, data and governance which will be best positioned to turn AI from a promising technology into a transformative force for better health outcomes.

For more information

See the WHO Europe report here. 

See the EHTEL Briefing paper here.

See background information on an earlier WHO Strategic Partners’ Initiative for Data and Digital Health (SPI-DDH) report here. 

See the work of the EvalCommunity Academy for monitoring and evaluation personnel here.  The academy has analysed the implications of the 2026 WHO report on AI training from a monitoring and evaluation perspective.