WEF: Why Is AI and Digital Healthcare Failing to Advance?
AI is already helping clinicians make better calls, automating paperwork, and speeding up discovery. Yet at a system level, progress feels slow. The World Economic Forum (WEF) flags a growing gap: ageing populations, higher demand, and a projected 10 million-worker shortfall by 2030, while costs outpace GDP. The need for change is obvious; scaling what works is the sticking point.
The promise is real. Adoption isn't.
WEF points to clear wins: more accurate diagnoses, stronger decision support, better remote monitoring, and a lighter administrative load. The American Medical Association notes over US$100bn has flowed into US digital health since 2010, but most tools never move past pilots. Over 70% of AI-related FDA approvals sit in imaging, with many still missing from day-to-day care.
FDA's AI/ML device list shows momentum, but it hasn't translated into routine clinical use at scale.
Why progress stalls
Healthcare isn't a single buyer. It's payers, providers, regulators, patients, and vendors operating under different rules across regions. Even strong point solutions hit a wall in this environment.
- Data is scattered, inconsistent, and hard to access for training models. Privacy and sharing rules add friction.
- Regulation is stringent and variable. Approvals don't guarantee reimbursement or clinical uptake.
- Markets are fragmented, making expansion beyond a pilot site slow and expensive.
- Human factors matter: burned-out staff, limited change capacity, and a bias to build in-house.
- Funding models are complex, with unclear incentives and payment pathways for digital care.
What leaders are saying
Thomas Buberl, Group CEO of AXA: "Business and political leaders need to move beyond a cost-savings approach toward imagining new models for growth and to guide citizens through a responsible deployment of AI. If AI is combined with prevention, it can help us stop paying for yesterday's damage and start preventing tomorrow's, making the system ultimately sustainable."
Dr Jochen Malinowski, Accenture: "AI can transform healthcare, if we fix the data foundation... remarkable potential in diagnostics, prevention and clinical workflows."
What works: the ARC model
One model getting traction is ARC (Accelerate, Redesign, Collaborate), launched in 2019 at Sheba Medical Centre in Israel. ARC connects health systems, industry partners, and 100+ startups to tackle workforce shortages, inefficiencies, burnout, and regulatory hurdles.
- Local-first integration: Embed digital tools into real clinical pathways, measure outcomes, iterate fast.
- Public-private partnerships: Share risks and benefits, align incentives early, and de-risk adoption.
- Global scaling platforms: Once a model works locally, scale it through a vetted network instead of starting from scratch each time.
Culture is make-or-break. ARC invests in curiosity, inclusivity, and transparent leadership. Through entrepreneurship training, mentorship, and design thinking, hospitals shift from passive adopters to active builders of solutions that fit their context.
Five moves health leaders can make now
- Fix the data layer: Prioritize interoperability, consent management, and secure access to anonymized datasets.
- Design for reimbursement: Involve payers early. Map billing pathways and clinical evidence needs up front.
- Start where pain is highest: Pick use cases tied to staffing shortages, throughput, or safety. Prove value in weeks, not years.
- Co-create with clinicians: Build workflow-native experiences. Reduce clicks, don't add them.
- Adopt through partnerships: Use shared-risk contracts and common evaluation frameworks to speed scale across sites.
Coordination at system level
This won't move without aligned effort. The WEF's Digital Healthcare Transformation Initiative aims to spread models that actually scale. Governments, investors, and health systems need to pull in the same direction to build more integrated, tech-enabled care.
If your teams need practical skills to evaluate, procure, and implement AI tools, consider focused training paths by role: AI courses by job. Upskilled clinicians and managers make adoption faster and safer.
Bottom line
AI isn't the issue. Fragmentation is. Build on solid data, tie projects to reimbursement and workforce relief, and scale through partnership. Do that, and digital health starts paying dividends where it counts: outcomes, access, and cost.
Your membership also unlocks:
