Inside an AI-enabled health care facility: Korea’s Samsung Medical Centre

The 10-Year Health Plan promises no less than to ‘make the NHS the most AI-enabled care system in the world’, acknowledging that the NHS is currently a ‘technological laggard’. But what does a genuinely AI-enabled health care system look like, and which systems can the UK learn from to accelerate this journey?

We are currently researching five hospitals across Singapore, the US, Germany, Canada, and South Korea where AI has been embedded at scale to improve the quality of patient care, staff experience and productivity. This blog shares some early insights from Samsung Medical Centre (SMC), a tertiary hospital in Seoul, South Korea, including from interviews with key hospital leaders involved in their work on AI.

AI use at Samsung Medical Centre

SMC is widely acknowledged to be a highly digitally mature hospital, with strong foundations in place that most AI tools depend on. Data are captured at every point of the patient journey, flowing through a single electronic patient record that is connected to every device in the hospital. These data are used not just for patient interactions but also for research and service improvement. 

AI is used across a wide range of pathways at SMC. For example, clinical AI solutions are used in pressure ulcer management, risk prediction and robotic assistants. AI is also used across non-clinical areas, including to schedule procedures and through SMC-GPT – the hospital’s in-house generative AI platform for medical record summarisation, document drafting, hospital protocol guidance and workplace queries. 

This breadth of use is the result of a sustained focus on building both a culture and governance structures that enable widespread AI adoption. 

An ‘arduous changing process’

While SMC’s AI use is advanced, the journey was not always easy. Wonchul Cha, Chief Information Officer at SMC, stressed that in Korea, ‘transformation’ is understood as Hyeokshin (革新), which Cha said can be translated as a ‘very arduous changing process’. 

Nearly two decades ago, when SMC proposed replacing paper records with online versions, many staff in the records department strongly objected. Several wrote to the hospital president to warn that the change was too risky with the technology available. Jong-Soo Choi, SMC’s current Chief Technology Officer, attributes this resistance to a gap in understanding rather than mere opposition to change. Today, after experiencing many successful innovations, the very same staff often lead even more ambitious projects. 

More recently, the introduction of a bedside AI image-capture tool for wound records posed some implementation challenges. Before using the tool, nurses had to undergo training to take images correctly and understand how each capture would be logged in the patient record. Adoption required months of iteration and a deliberately staged rollout. It took several incremental phases starting with two teams and doubling until it reached the whole hospital.

A culture of push and pull  

SMC’s leaders aim to create a culture of AI transformation – not limited to protocol and education. This is fostered through continual improvement, trust and peer support. An anonymous bulletin board receives daily suggestions from staff, ranging from changing hospital signage to adapting a workflow to investing in a particular technology. This bulletin is monitored in real time by the hospital leadership. Any tool or modification that has potential to improve the quality of care is taken under consideration. 

SMC’s AI governance structure is designed to enable a bi-directional system: top down and bottom up. Staff are encouraged to suggest problems that need solving and consider the potential of AI to do so. At the same time, SMC’s leadership push for institution-wide AI implementation, including non-clinical use cases. With staff advocating for AI tools rather than passively adopting them, SMC can find where AI is most needed and how it might be most effectively implemented. These bi-directional initiatives also enable a transparent assessment of whether the solution was effective.

The AI governance structure is comprised of both senior leaders and multidisciplinary clinical teams. There is an executive council – the AI Strategy Board – that oversees AI transformation, and a cross-functional AI Working Group supports all aspects of transformation. SMC’s Office of AI Transformation (AX) pulls engineers and coordinators together to effectively execute decisions from end to end. This structure is flexible, with SMC adapting as needed to keep up with the changing nature of AI.  

The national context

Korea’s national policy context, alongside critical research and seed funding from government, provides clarity and support for SMC’s AI use. Korea’s landmark AI regulation came into effect in January 2026 with the AI Basic Act, enacted by a National AI Committee (chaired by the president) and supported by a network of advisory subcommittees, including one dedicated to health care. Korea is one of only seven OECD countries with a published national health AI strategy. 

However, the direction was not always clear from the start. Meong Hi Son, Chief Medical Information Officer at SMC, described a familiar bind: regulators require evidence of safety and effectiveness before approval, but much of this evidence can only be generated once a tool is in use. While the government has become more willing to allow exploratory studies, changing guidance can sometimes cause confusion. This challenge is not unique to Korea. The UK’s National Commission into the Regulation of AI in Healthcare, which published its recommendations this month, found the current framework ill-suited to AI systems. It proposed staged authorisations to allow supervised deployment while evidence is gathered. 

At the moment, there is no playbook for how a health care provider should become AI-enabled. Health care systems across the world are figuring it out in real time, with different organisations taking varied approaches. Our full report, to be published next year, will outline five examples from health care systems leading the way.

Alex is hosting a panel featuring Dr Wonchul Cha, Samsung Medical Centre's Chief Information Officer, at the Health Foundation’s annual AI in the NHS event on Thursday 15 October. Register to join online.