From NHS Pilot to Practice: Closing the Gap Between AI Innovation and Deployment

Rt Hon Andrew Stephenson CBE

Chair, Curia, Health, Care and Life Sciences Research Group

Following the publication of AI Is Ready. Is the System? report by Curia and UKAI, Chair of Curia’s Health, Care and Life Sciences Research Group and former Minister of State at the Department of Health and Social Care, Andrew Stephenson writes that the NHS is well positioned to lead in AI-driven healthcare, but the challenge has shifted from innovation to implementation.

Across government, the NHS, academia, and industry, there is broad agreement that AI and advanced data science will transform healthcare and life sciences, particularly within the NHS. The science is advancing rapidly. British universities are among the strongest in the world. Our life sciences sector contributes more than £100 billion to the economy and supports hundreds of thousands of skilled jobs.[i] We have longitudinal health records, high GP registration, and a unique ability to link data across the life course. In many respects, we are better positioned than almost any nation, especially considering the advancements in the NHS.

And yet, as these discussions make clear, the central challenge is no longer invention. It is implementation.

From Innovation to Implementation: Why the NHS Must Move Beyond Pilots

Too often, innovation in the NHS remains confined to pilots. Promising technologies are tested, evaluated and praised – and then stall before reaching routine practice. The reasons are not mysterious. They include fragmented procurement, unclear lines of accountability, data governance complexity, workforce pressures, and, at times, a cultural instinct to equate the status quo with safety.

We must challenge that assumption directly. Doing nothing is not risk-free. In a health system and NHS facing rising demand, workforce shortages, and widening inequalities, standing still carries consequences of its own. Innovation, when properly evaluated and safely deployed, is not a threat to patient safety – it is an essential route to improving it.

Aligning NHS Regulation, Infrastructure and Growth to Deliver AI at Scale

Throughout these roundtables, three themes emerged repeatedly.

First, system readiness matters as much as technological capability. AI tools do not deploy themselves. They depend on infrastructure, workforce confidence, digital maturity, and governance that is proportionate rather than paralysing. Building the bridge from innovation to adoption requires effort on both sides.

Second, regulatory reform must strike the right balance. We must be safe, but we must also be fast and trusted. The regulatory framework for medical devices was not designed with adaptive AI in mind. We therefore need approaches that allow continuous monitoring, post-market learning, and clear accountability, without creating unnecessary delay. Businesses need clarity, boards need assurance, and patients need confidence.

Third, we must align economic growth with NHS transformation. Startups and scaleups cannot wait two years for data access while their runway expires. Equally, trusts cannot be expected to underwrite infrastructure without national support. If we want Britain to retain sovereign capability in AI-enabled healthcare, we must ensure that procurement pathways, data environments, and funding models enable responsible domestic innovation to scale.

Prevention and predictive medicine are particularly instructive. The science in genomics and risk stratification is advancing rapidly. Yet our funding structures in the NHS remain weighted toward treating illness rather than anticipating it. If we are serious about shifting from reactive to preventative healthcare, budgetary and accountability frameworks must evolve accordingly.

As Chair of Curia’s Health, Care and Life Sciences Research Group, my focus is on turning policy into practice. It is not enough to announce strategies. We must translate ambition into delivery at trust, system, and national level.

The discussions summarised in this report are candid – but practical. They highlight barriers to the adoption of AI in the NHS, but they also demonstrate appetite for change. Clinicians want tools that give them more time with patients. Innovators want clarity and partnership. Policymakers want solutions that improve outcomes and support growth.

Our task now is to convert shared insight into coordinated action. If we do so, the UK can become not just a leader in AI research, but a leader in responsible, system-wide deployment – improving patient care, strengthening our healthcare and life sciences sectors, and ensuring the NHS remains sustainable for generations to come.

References

[i] https://www.gov.uk/government/news/life-sciences-sector-plan-to-grow-economy-and-transform-nhs