How to get hospital AI from pilot to scale (Eyal ZImlichman, Sheba Medical Centre)

Three in ten AI pilots at Sheba Medical Center never scale. Eyal Zimlichman on why, and what Europe's regulation holds back. Nine years after his first Faces of Digital Health interview, when ARC was still being built inside Sheba Medical Center, Eyal Zimlichman returns with the numbers. He explains how Sheba measures the return on AI (clinical outcomes, cost and clinician burnout), why its chief quality officer rather than its AI officer signs off every deployment, and why about 30% of pilots are discontinued. We discuss what ambient scribes change for patient safety, documentation and malpractice, why hospitals still treat start-ups as a nuisance, why no European hospital belongs to Mayo Clinic Platform's federated data network, and how ARC plans to grow tenfold in 2027. GUEST Prof. Eyal Zimlichman — Chief Transformation, Innovation and AI Officer, Sheba Medical Center, Israel; founder and director, ARC Innovation Host: Tjaša Zajc WHAT THE CONVERSATION COVERS \- What has changed at Sheba Medical Center and ARC since 2017 \- How a hospital measures AI return on investment: outcomes, cost, access and burnout \- Aidoc and AI triage for intracranial haemorrhage: the Sheba mortality study \- AI governance in hospitals: why the chief quality officer approves AI deployments \- Time-limited AI approvals: 3–12 month reviews based on risk \- Ambient scribes: accuracy differences between vendors and medication-name errors \- Ambient documentation, billing and the case for fewer malpractice claims \- Why 30% of hospital AI pilots fail to scale, including an internally built discharge alert \- Open innovation: why healthcare systems distrust start-ups \- Procurement, tenders and the barriers to bringing in new technology \- Europe as a "regulation superpower": unclear rules, cautious lawyers and the Charité ARC centre \- Mayo Clinic Platform and federated health data sharing without European partners \- Hospitals as economic engines: ARC start-ups, jobs and KPIs \- China, AI regulation and the speed of adoption \- ARC Global in 2027: tenfold scale-up and AI-native hospitals by 2030 CHAPTERS 02:43 Nine years on: Sheba, ARC and AI since 2017 06:52 How to measure AI return on investment in a hospital 10:58 Patient safety and ambient scribes: who approves AI 15:51 Ambient documentation, billing and malpractice 18:06 The pilots that didn't scale, and why 21:47 Open innovation: why hospitals distrust start-ups 26:35 Europe as a "regulation superpower" 30:43 A hospital as an economic engine: start-ups and jobs 36:13 China, AI regulation and speed 39:27 ARC in 2027: tenfold growth and AI-native hospitals MENTIONED ARC Innovation at Sheba Medical Center • ARC Global • Charité BIH ARC Innovation Center (Berlin) Aidoc • Mayo Clinic Platform • ARC Summit, Tel Aviv, 22 October 2026 Orion et al., International Journal of Emergency Medicine (2023) — AI triage for intracranial haemorrhage Earlier FoDH episode: Eyal Zimlichman, 2017 \[LINK] FACES OF DIGITAL HEALTH Website: https://www.facesofdigitalhealth.com Newsletter: https://fodh.substack.com LinkedIn: https://www.linkedin.com/company/faces-of-digital-health Spotify: https://open.spotify.com/show/4cElKJHrauyP6QJQaCkvdY Apple Podcasts: https://podcasts.apple.com/gb/podcast/faces-of-digital-health/id1194284040 #healthcareAI #digitalhealth #hospitalinnovation #ambientAI #AIgovernance #patientsafety #ShebaMedicalCenter #healthtech #openinnovation #healthpolicy #AIregulation #FacesOfDigitalHealth