“That's what EHRs were supposed to do too, and that never happened.”
Session 2 brought together senior women from across digital health, medtech, pharma and healthcare technology for a wide-ranging and genuinely candid conversation. The theme of AI adoption quickly expanded into something broader: the gap between what the industry is promising and what is actually being experienced on the ground.
1. AI literacy: the constraint nobody's talking about loudly enough
The pace at which any organisation can truly integrate AI is largely determined by its overall literacy, starting at board level. Where boards lack understanding, fear tends to fill the gap, and that fear cascades downward, slowing progress at every level.
The comparison to sustainability transformations resonated: a similar arc of hype, fear, pockets of expertise and uneven integration.
Even organisations with genuine AI expertise on their boards are struggling to get adoption to trickle down. One approach shared: give high performers who already embrace AI a visible platform internally, letting peers drive change rather than leadership mandating it. The word “augment” came up repeatedly as a deliberate reframe away from replacement.
2. The implementation gap
The reality of AI implementation is consistently messier than the pitch. Nurses in one pilot programme responded bluntly: “That's what EHRs were supposed to do too, and that never happened.” The promise of time savings has been made before and not delivered, and that history is creating real scepticism on the clinical frontline.
- Lack of clear problem definition: tools arriving before the problem is understood
- No success metrics: not knowing how to measure whether AI is working
- ROI pressure: demand for returns before solutions have time to embed
- Speed of release: tools reaching market faster than organisations can evaluate them
The consensus: the most successful AI implementations are narrow, well-defined and focused on removing specific friction, rather than transforming entire workflows overnight.
3. AI and the patient: starting in the wrong place
Too much innovation is designed around the system rather than the patient. Because the money lives in pharma and large health systems, solutions are often built for those buyers rather than the people at the end of the journey.
What was championed instead: starting from the patient's daily reality. AI has a genuine opportunity to personalise journeys in low-risk, immediately valuable ways, from pre-diagnosis engagement to reducing the friction of navigating care. Preventative care remains underfunded on both sides of the Atlantic, and AI may be one of the more practical tools for shifting that.
4. Europe vs US: more nuanced than the headlines
Is Europe ahead of the US in digital health? It depends entirely on the sector. One participant described a UK health tech conference where vendors showcased cutting-edge AI and robotics while NHS leaders in the same building presented on their struggles to implement basic EPR systems. The more useful question is where specific use cases are being implemented well, and what is enabling that.
5. Governance: solving the right problem
How do you make sure the AI being pursued solves a real problem rather than being adopted because it is exciting? Start with the problem, not the technology.
Reimbursement was flagged as a structural blocker governance alone cannot fix. In medtech and hospital settings especially, even a solution that clearly adds value struggles without a reimbursement pathway: an area where political pressure is needed alongside commercial innovation.
6. Cybersecurity: the parallel conversation
As AI becomes embedded in healthcare workflows, cybersecurity is an increasingly urgent concern, particularly when building secure AI platforms in regulated clinical environments.
7. The literacy gap is showing up everywhere
Including in the talent market. Demand is growing for people who can bridge clinical understanding and AI implementation, and that profile remains genuinely rare.
Resource shared
Book: The AI-Driven Leader, a practical, accessible read on implementation strategy (also on audio).
All contributions are anonymised. Nothing shared is attributed to any individual.
