Q&A: AWS’ Sasha Rubel on AI Regulation & European Healthcare

Sasha Rubel leads AI Policy across Europe, the Middle East and Africa at Amazon Web Services (AWS).

Working with governments to support their understanding of what AI means in practice and helping technology leaders navigate the rules shaping how AI is built and deployed, Sasha is on the frontline of AI policy across the EMEA region.

Before joining AWS, Sasha gained deep experience in international affairs, spending over a decade at UNESCO, including positions in Ethiopia and Senegal.

Following her time on the African continent, Sasha returned to France with UNESCO to lead the organisation’s global work across AI, innovation, data governance, digital inclusion and digital transformation.

Today, a significant portion of her work involves advising governments and technology leaders about AI regulation and deployment in the healthcare sector, where “the consequences of getting AI right – or wrong – can be profound,” says Sasha.

Sasha shared her insights with Healthcare Digital on the state of play across the European regulatory landscape and where the opportunities and barriers for widespread AI adoption in healthcare lie.

What is Europe getting right on AI adoption in healthcare policy - and is the balance wrong anywhere?

SR: Our new data from an AWS-commissioned report, ‘Unlocking Europe’s AI Potential’, based on a survey of more than 34,000 European organisations, businesses, and citizens, shows incredible momentum and opportunity for AI adoption in the healthcare space.

AI adoption in healthcare is still below the European business average - 41% compared with 54%. But among healthcare organisations that are adopting AI, use is more sophisticated than in any other sector: 55% are consistently using multiple AI tools. Healthcare is also leading on AI governance, with 36% of organisations having a formal AI strategy.

Taken together, this suggests a more deliberate approach to adoption: healthcare organisations may be moving more slowly, but those that do adopt AI are putting the governance and strategy in place alongside deployment, rather than deploying first and addressing the risks later. In a sector where patient safety and clinical trust are at stake, that disciplined approach matters.

Where the balancing act becomes challenging isn’t this caution itself, but the complex regulatory environment innovators have to navigate. For example, a clinical AI tool supporting diagnostic decisions is faced with answering several differing compliance regimes, all at the same time. 

It may need to obtain CE marking under the medical device framework, ensure that any processing of patient data complies with the General Data Protection Regulation (GDPR), and meet the requirements that apply to high-risk AI systems under the EU AI Act. Each has its own authority, compliance obligations and enforcement path. 

Over half of healthcare organisations say that compliance spend has risen significantly – more than any sector – and that spend is going toward reconciling overlapping rules rather than towards life-saving innovation and deployment.

Europe hasn’t matched its adoption rates with an equally coordinated and mainstreamed regulatory pathway, and that gap is absorbing resources and momentum that would otherwise translate directly into patient impact.

How important is a joined-up approach to AI regulation, and what would that look like in practice?

SR: It’s absolutely crucial. In practice, a joined-up approach means making it as simple as possible for organisations to understand and comply with the rules: one coherent framework, common definitions, and a clear point of contact for regulators across the EU. 

Today, a founder can face different rules, different interpretations, and different regulatory bodies across 27 countries. That creates unnecessary complexity and can slow innovation. There are positive signs that this is being recognised. 

The Commission’s simplification proposals are a step in the right direction - the challenge now is turning that ambition into a genuinely consistent approach across Europe.

What needs to change to move from successful pilots to widespread adoption?

SR: Moving from successful pilots to widespread AI adoption in healthcare requires Europe to address the structural challenges that prevent promising AI solutions made in Europe from scaling. 

Regulation and enforcement need to be simplified, harmonised, and predictable across the EU single market and build on international standards to facilitate companies scaling. 

Secondly, Europe needs to treat AI skills as key infrastructure, giving leaders, the workforce, and citizens the confidence to integrate AI into core workflows rather than leaving it as a side project. 

Finally, Europe needs stronger routes to growth capital, procurement and international markets, alongside closer collaboration between government, industry and civil society. 

The goal is not deregulation, but clear, simplified, and trusted rules and the conditions to turn innovation into sustained, measurable outcomes.

Are there European AI companies you think could enact real transformative change in healthcare?

SR: Europe has several AI companies that could drive genuinely transformative change in healthcare because they are addressing real clinical and patient-access challenges. 

• Iktos, for example, combines generative AI with automated laboratory robotics to accelerate drug discovery, potentially reducing development timelines from years to months. 
• Proximie is tackling inequalities in surgical care by connecting operating rooms, allowing specialists to support and train colleagues remotely in real time. 
• myTomorrows uses AI to help patients and physicians navigate clinical trials and pre-approval treatments, improving access to therapies that might otherwise remain out of reach. 
• And Callyope is using AI to identify early signs of mental-health relapse between appointments, enabling more preventative intervention. 

Together, these companies demonstrate how Europe can translate AI, research and clinical expertise into measurable improvements in patient outcomes.

If you could change one thing about EU AI policy to support health startups scaling up, what would it be?

SR: If I could change one thing, it would be to make the rules simpler, clearer and more consistent across Europe. Not because regulation matters less, but because clarity gives innovators the confidence to move responsibly.

The opportunity is extraordinary. 39% of healthcare companies believe AI could fundamentally transform their industry. Imagine what that could mean: earlier diagnoses, faster drug discovery, more personalised care, and better outcomes for patients.

Europe has the talent, the research and the ambition to lead here. What we need is an environment that allows the people building these solutions to scale them. Ultimately, this isn't about AI. 

It's about what AI makes possible for people - finding cures faster, giving doctors better tools, and giving patients more hope. The real job now is to make sure the best ideas don't stay ideas. They reach the people who need them most.

AWS - key healthcare partners

Pfizer

CEO: Albert Bourla | HQ: New York

Pfizer and AWS collaborate through the Pfizer-Amazon Collaboration Team (PACT), applying cloud, AI, machine learning and analytics to drug development, clinical manufacturing and supply chains. The partnership has expanded into generative AI, with Pfizer using AWS technologies to develop prototypes and accelerate scientific workflows. AWS says the collaboration has saved scientists up to 16,000 hours of search time annually. 

Novo

CEO: Maziar Mike Doustdar | HQ: Bagsværd, Denmark

Novo Nordisk selected AWS as its strategic cloud and AI partner in August 2026, deepening a relationship focused on accelerating drug discovery and modernising operations. The partnership combines Novo Nordisk’s therapeutic expertise with AWS cloud and AI technologies, including Amazon Bedrock, to support research, development and enterprise transformation across the global healthcare company. 

Moderna

CEO: Stéphane Bancel | HQ: Cambridge, Massachusetts

Moderna has standardised its real-world data strategy on AWS, using AWS services to ingest, store and analyse large scientific datasets. The companies’ wider relationship spans manufacturing, research and data infrastructure, with AWS supporting Moderna’s mRNA platform and digital operations. An AWS case study reports that Moderna achieved 70% faster data extraction and analysis using AWS Data Exchange.