Healthcare Isn’t Europe’s Biggest AI Adopter, But It Is Seeing The Best Returns

When it comes to AI adoption in Europe, healthcare isn’t leading the pack.

According to Amazon Web Services’ latest “Unlocking Europe’s AI Potential” report, only 41% of healthcare organisations have adopted AI, compared to a European average of 54% across industries. Yet, among those that have gone ahead and embraced the technology, healthcare is outperforming almost everyone else. The report found that healthcare organisations are more likely to use multiple AI tools, more likely to have formal governance structures in place and they’re more likely to report strong returns on their investment.

For many, this may seem backwards, at least at first glance. After all, we’ve been told, ever since AI really took the world by storm, that AI is here to make our lives easier, remove tension, make our jobs more efficient and help us save money. But for quite some time, another school of thought has been that this may be true, but only if AI, much like any other new technology, is used and implemented properly, thoughtfully and effectively. It’s not a quick fix and it won’t automatically improve everything. And the results of this AWS report seem to demonstrate exactly that.

For the past two years, if not longer, AI discussions have largely been dominated by a race to adopt. Businesses have rushed to launch copilots, experiment with chatbots and announce AI strategies. In fact, for many businesses, AI adoption rates have been perceived as  a marker for innovation and success without actually evaluating the results of said AI use.

But the healthcare industry’s experience suggests that success may have less to do with how quickly organisations adopt AI and more to do with how deliberately they do it.

Slow Doesn’t Necessarily Mean Behind

The report seems to tell us a story about a sector that has been more cautious than most, and understandably so. Indeed, healthcare organisations are less likely than the European average to have adopted AI at all. However, among those that have, 55% consistently use multiple AI tools, compared to 45% across industries, and only 13% remain in the experimentation phase – thus, they progress beyond “just trying it out” to proper implementation.

Indeed, healthcare also leads the pack when it comes to AI governance, with 36% operating under a formal AI strategy and 18% having formal data governance frameworks, which according to AWS, is nearly double the cross-industry average.

The reasons for the slower pace and caution here are quite obvious. In retail, a faulty recommendation engine might annoy customers; in marketing, a hallucinated paragraph might create an embarrassing social media post; but in healthcare, mistakes can have far more serious consequences. Clinical decisions, diagnoses and patient records leave far less room for error.

As a result, healthcare organisations often have to answer difficult questions way before deployment is even on the table. They need to seriously consider things like whether or not the data is reliable, whether decisions made can be explained and justified, accountability for actions when things go wrong and how patient information will realistically be protected.

Many business leaders loathe questions like these because they’re difficult to answer and end up making adoption, and inadvertently, a little bit less exciting. But, the report suggests going slow may also improve outcomes.

AI Is Solving Real Problems…Just a Little Slower

Another possible explanation lies in how healthcare is using AI. Many organisations across Europe are still using AI for relatively basic tasks. AWS’ wider research found that while AI adoption has accelerated rapidly, only a minority of businesses have reached what it classifies as advanced adoption, where AI becomes embedded in core processes rather than sitting on the edges of the business.

Healthcare, however, appears to be moving in that direction faster than many sectors.

The report identifies diagnostics, operational efficiency and clinical decision support as leading priorities. These are high-value problems that are tied directly to patient outcomes, clinician productivity and healthcare capacity.

The examples highlighted by AWS reinforce that point. Munich Leukaemia Laboratory is using AI and cloud-scale genomics to identify rare forms of leukaemia in hours or days rather than weeks. Iktos is combining AI with laboratory robotics to accelerate drug discovery. Callyope is developing systems that can identify early signs of mental health relapse, while Proximie is using software and AI-powered technologies to help improve access to surgical expertise.

These are projects with targeted applications that are designed to solve existing challenges.

Believe It Or Not, Better Governance May Be Producing Better Returns

The healthcare sector’s stronger returns may also be linked to governance. The report found that 28% of healthcare organisations said AI returns significantly or massively exceeded their investment, while 60% reported increased revenue and 55% reported major productivity gains, all of which were well above the cross-industry average.

Indeed, it would be easy to assume that faster adoption automatically produces better results, and that certainly has been the most common belief for quite some time. But while many people have been asserting otherwise, we now have some data to back the claims.

Across Europe (and the rest of the world), businesses have been adopting AI rapidly, but many are still struggling to move beyond pilots and experiments. They’ve been under pressure, both within the industry and specific businesses too, to keep up and throw AI at everything. However, AWS’ broader research found that only 22% of organisations have reached advanced AI adoption, despite more than half now using AI in some form. Thus, the old-school “move fast and break things” mentality simply hasn’t been effective when it comes to AI.

The healthcare industry‘s approach has been quite different, in that rather than prioritising widespread deployment, many organisations seem to be focusing on governance, data quality and clear use cases before scaling. Evaluate, prepare and implement properly, in order to be efficient and reap long-term benefits beyond the mere excitement of AI.

It may not be the most flashy approach, but it certainly seems to be one of the main reasons that the sector is seeing stronger returns than other sectors.

The Cost Of Being Careful

That isn’t to say healthcare has solved AI adoption or is doing things perfectly. In fact, the report identifies some of the sector’s biggest challenges, even in the context of being potentially more successful than other industries.

Only 24% of healthcare organisations rate their workforce’s digital skills as good or excellent, compared to 36% across industries, and more than half cite skills shortages as the biggest barrier to AI adoption. At the same time, 56% report significantly higher compliance costs, reflecting the realities of operating in one of Europe’s most heavily regulated sectors.

So, healthcare may be generating strong results, but it is doing so while carrying a heavier burden than most industries. And in many ways, this almost enforces the approach that the industry has taken and shows that perhaps we could be even more cautious and measured. Now, the sector’s challenge now is whether it can maintain that balance: too much caution could slow innovation, but too little could undermine trust. So, will the industry have enough support to be able to continue to be successful?

What Can Other Industries Learn from the Healthcare Sector’s Approach?

The approach taken by the healthcare industry isn’t a blueprint for success, but it still holds lessons that others may be able to learn from. The biggest and most important of which may be that AI success isn’t (and shouldn’t) necessarily be measured by adoption rates, because that is not a measure of success in itself.

After all, healthcare isn’t Europe’s biggest AI adopter, but it is among the sectors seeing the strongest outcomes. And that suggests that organisations may need to focus less on whether they’ve adopted AI and more on how they’re adopting it and why.

The businesses that are generating the greatest value from AI are often not those deploying the most tool; it’s the ones that are embedding AI into meaningful workflows, investing in governance and solving clearly defined problems.

For a sector where the stakes are measured in patient outcomes rather than productivity metrics alone, healthcare has had little choice but to be careful. And caution, it turns out, may be the core ingredient for success.