Already in use

Healthcare is a hard test of whether a capable system becomes a working one. The stakes are clinical, the data are sensitive, and the workflow is already full.

WHO Europe’s assessment of AI readiness, covering its 53 member states, shows the gap in plain figures. Nearly two thirds of countries are already using AI in diagnostics. Half have introduced AI patient chatbots. Only 8% have a health-specific AI strategy, and only 8% have liability standards for when a system fails. One in five countries provides AI education before a health professional qualifies. One in four offers training once they are in the workforce. Fewer than half have assessed whether their legal frameworks are fit for the technology, and almost 40% have no ethical guidance on AI in health.

What the figures are actually measuring

Those are not model benchmarks. They are questions of responsibility, data, workforce and how a tool sits inside a service that already has a way of working. A system can perform well in a trial and still fail if the record it needs is elsewhere, if nobody is accountable for an error, or if staff have not been trained to question it.

The same pattern shows up beyond diagnostics. Administrative tools, imaging, research software and operational systems all have to pass procurement, safety review, interoperability and the cost of changing the surrounding process. Technical performance is the start of that list, not the end.

WHO’s own statement puts the priority in that order: governance has to keep pace with deployment, including strategy, liability and training for the people who use the tools. Through 2027, the useful signal in healthcare AI is not a new demonstration. It is whether systems are being used, safely and repeatedly, inside real services.

Sources

WHO Regional Office for Europe, “Govern AI in health before the gaps become irreversible”, 15 July 2026. Health-specific AI strategy in 8% of countries in the WHO European Region. Nearly two thirds deploying AI in diagnostics. Half using AI patient chatbots. One in five provide pre-qualification AI education. One in four offer in-service training. Fewer than half have assessed legal frameworks. Almost 40% have no ethical guidance.
https://www.who.int/europe/news/item/15-07-2026-statement---govern-ai-in-health-before-the-gaps-become-irreversible

WHO Regional Office for Europe, Lisbon meeting note, 15 July 2026. Only 8% have liability standards defining responsibility when an AI system fails.
https://www.who.int/europe/news/item/15-07-2026-who-brings-37-countries-together-in-lisbon-to-get-ai-governance-right-and-make-it-work-for-every-patient

Stone Maven, Markets and Focus 2026 to 2027, public outlook, September 2026, Healthcare, pp. 29 to 32. The report frames health AI through strategy, law and ethics, data, workforce and delivery, citing WHO Europe’s assessment of EU health systems. The figures in this note are the later Region-wide statement, which is wider than the EU-only assessment.
https://docs.stonemaven.co.uk/markets-focus-2026-2027.pdf