AI in Healthcare Delivers Real Results, but WHO Says Better Policies Will Decide Its Global Success

WHO's latest report shows that AI is already improving diagnosis, hospital efficiency and patient care across seven countries, but its success depends on strong governance, high-quality data, skilled health workers and continuous monitoring rather than technology alone. The report urges governments, development partners and private-sector stakeholders to invest in digital infrastructure, ethical AI regulation and workforce capacity to responsibly scale AI and strengthen health systems amid a projected global shortage of 11 million health workers by 2030.

Artificial intelligence is rapidly moving from experimental projects into everyday health-care delivery, but the World Health Organization (WHO) Regional Office for Europe says technology alone will not solve the growing pressures facing health systems. In its latest report developed under the Strategic Partners' Initiative for Data and Digital Health (SPI-DDH), WHO presents evidence from 11 real-world AI implementations across Finland, Israel, Latvia, Norway, Portugal, Spain and the United Kingdom, showing that successful AI adoption depends on strong governance, reliable health data, skilled professionals and continuous monitoring rather than sophisticated algorithms alone. With the world expected to face a shortage of 11 million health workers by 2030, the report argues that AI can help health systems improve productivity, reduce administrative workloads and strengthen clinical decision-making without replacing health professionals.

AI Is Already Delivering Measurable Results

The report highlights several examples where AI has produced measurable improvements in clinical care and hospital operations. Helsinki University Hospital in Finland reported a 25% increase in the detection of incidental pulmonary embolism among cancer patients, allowing treatment to begin approximately two weeks earlier. Norway's Haukeland University Hospital achieved up to 80% reductions in organ-contouring time for head and neck radiotherapy planning and around 50% savings in breast and prostate cancer cases, helping clinicians deliver faster and more consistent cancer treatment.

Israel demonstrated some of the strongest clinical outcomes. At Sheba Medical Center, AI-assisted cardiac ultrasound supported bedside diagnosis for 660 patients, influencing clinical decisions in 55% of cases, including treatment adjustments, early discharge and improved patient management. Clalit Health Services also used machine learning to identify undiagnosed hepatitis C patients. Instead of screening 50,000 people to detect only a few dozen cases, the AI system identified 500 high-risk individuals and discovered 38 positive patients, representing nearly a 100-fold improvement in screening efficiency.

Operational gains were equally significant. In Spain, AI-based transcription of medical device readings reduced administrative work by 123.74 staff hours per month at each long-term care facility while achieving staff satisfaction scores of 8.9 out of 10. In the United Kingdom, agentic AI deployed in emergency departments reduced reporting time for senior managers by 50% and administrative workloads by 30%, allowing more time to be devoted to patient care.

Better Policies Will Decide Whether AI Succeeds

The report makes it clear that AI implementation is primarily a governance challenge rather than a technology challenge. Every successful project relied on high-quality health data, strong cybersecurity, interoperable digital systems and close collaboration between clinicians, data scientists, IT specialists and regulators. Organizations that treated AI as a long-term transformation rather than a one-time software purchase achieved the best outcomes.

WHO highlights several international frameworks supporting responsible AI adoption, including the European Union Artificial Intelligence Act, the European Health Data Space, WHO's ethical guidance and OECD recommendations. Together, these frameworks promote transparency, accountability, privacy protection, human oversight and continuous evaluation throughout an AI system's lifecycle.

For governments, the message is straightforward: investment should extend beyond AI software to include digital health infrastructure, electronic health records, workforce training, cybersecurity and national governance systems capable of monitoring AI performance after deployment. Without these foundations, even highly advanced AI systems may fail to deliver reliable clinical benefits.

Why This Matters for Development Partners and Private Investors

The report carries important implications for development agencies, multilateral banks and private-sector technology companies. International development partners have an opportunity to accelerate responsible AI adoption by financing digital infrastructure, strengthening national health information systems, supporting regulatory reforms and improving AI skills among health workers. WHO also recommends creating regulatory "sandboxes" where governments, hospitals and technology developers can safely test AI solutions before introducing them into national health systems.

For private-sector companies, demand for AI-enabled diagnostics, workflow automation and predictive analytics is expected to grow rapidly. However, vendors will face increasing expectations to provide more than powerful algorithms. Hospitals now require technologies that integrate smoothly with existing workflows, comply with evolving regulations, protect patient data and demonstrate measurable clinical and economic benefits. Companies capable of supporting long-term implementation, monitoring and continuous improvement will have a stronger competitive advantage as governments introduce stricter AI governance standards.

A Practical Roadmap for Scaling AI Responsibly

The WHO report concludes that countries should adopt AI through a phased approach based on their institutional capacity. The first priority should be improving health data quality, interoperability and workforce AI literacy while establishing ethical safeguards. The next stage involves launching pilot projects, innovation hubs and monitoring systems to evaluate AI performance under real clinical conditions. Countries with mature digital health systems can then scale successful pilots nationally and develop comprehensive AI policies supported by real-world evidence.

The report ultimately shows that AI is already improving diagnostics, reducing administrative burdens and helping clinicians make faster decisions. Yet its long-term success will depend on sustained investment in digital infrastructure, trusted governance, workforce development and international collaboration. For policymakers, development partners and private-sector stakeholders, the evidence is increasingly clear: artificial intelligence is becoming a strategic pillar of modern health systems, but responsible implementation, not technology alone will determine whether it delivers lasting improvements in health outcomes, economic efficiency and equitable access to care.