AI in Health Systems: Big Potential, Slow Adoption and the Road to Scale

Artificial intelligence is already part of modern healthcare. It helps hospitals manage records, supports doctors in diagnosis, and improves efficiency behind the scenes. But despite this progress, AI has not yet reached its full potential.

A new OECD report, developed in collaboration with organisations such as the World Health Organization, the Global Digital Health Partnership, and the Coalition for Health AI, explains why. The technology is advancing, but its use across entire health systems remains limited and uneven.

In many countries, AI is widely used for administrative tasks. However, more advanced uses, such as medical imaging or predictive care, are still largely in pilot stages. Only a small number of these tools have been scaled across national healthcare systems.

The Big Challenge: Balancing Speed and Safety

Healthcare is different from other industries. It cannot move fast at the cost of safety. Every tool must be tested carefully because mistakes can affect lives.

This creates a difficult balance. Move too fast, and there is risk. Move too slowly, and patients miss out on better care. The OECD calls this challenge "responsible scale" – growing AI use while keeping patients safe.

The report warns that waiting too long can also be harmful. AI can improve early diagnosis, reduce workload for doctors, and make systems more efficient. Delays mean losing these benefits.

Data Problems Are Slowing Everything Down

One of the biggest barriers is data. Healthcare produces huge amounts of information, but much of it is not usable for AI.

Data is often stored in separate systems that do not connect. In some cases, it is incomplete or not standardised. Without clean, connected, and reliable data, AI cannot work properly.

Some countries are improving this by creating national data systems and better rules for sharing information safely. But progress is uneven, and many systems are still fragmented.

People, Trust, and Skills Matter Just as Much

Technology alone is not enough. Doctors, nurses, and healthcare workers need to understand how to use AI tools. They must be able to trust the results and know when to question them.

This means investing in training and education. Some countries are already adding AI learning into medical programmes and professional training.

Public trust is also critical. Patients need to feel confident that their data is safe and that AI is being used responsibly. Clear communication and transparency can help build that trust.

Rules and Leadership Make the Difference

The report highlights that countries with clear strategies and strong leadership are moving faster. Governments that have national AI plans, clear regulations, and testing environments for new technologies are better prepared.

Regulatory sandboxes, for example, allow new AI tools to be tested safely before full rollout. Updated approval systems and procurement rules also help bring innovation into healthcare faster.

However, not all countries have these systems in place. This creates gaps in progress and limits how widely AI can be used.

A Global Effort Is Needed

AI in healthcare is not just a national issue. The report stresses the importance of countries working together. Shared standards, common rules, and cooperation can help speed up progress.

Without coordination, systems remain fragmented, and innovation slows down. With it, AI can scale more quickly and benefit more people.

The message is clear. AI has the power to transform healthcare, but only if it is implemented carefully and at scale. The future will depend not just on new technologies, but on how well countries manage, trust, and work together to use them.