How hospitals manage the roll-out of AI products: Insights from global leaders
On 1 April, the IHF’s Harnessing Big Data Special Interest Group brought together a panel of international experts to explore how healthcare leaders are navigating the evolving landscape of artificial intelligence and big data. The panel shared their experiences of managing the roll-out of AI products in their hospitals, and discussed some of the challenges they have encountered as well as potential solutions.
Dr David Levine, Senior Vice President and Chief Medical Officer at Vizient (the SIG sponsor based in the USA), chaired the session and began by framing the discussion with insights into how AI is rapidly evolving within the US healthcare market. With AI adoption now reaching over 89% of providers in some capacity, the challenge is no longer about access to technology, but rather about strategic leadership and implementation.
“Healthcare is lagging behind other industries in adopting AI. This is mostly due to safety concerns and the lack of clear return of investment.” – Dr David Levine
However, healthcare executives are beginning to use AI to improve efficiency, incorporating tools supporting workstreams from documentation automation to patient risk stratification, with promising results. Importantly, he noted that AI should always support clinicians, not replace them, helping them to become more efficient or providing higher acuity.
Dr Gal Goshen, Chief AI Officer at ARC Innovation, Sheba Hospital (Israel), emphasised how innovation needs structure and strategy, embedded in a culture of continuous improvement. At Sheba’s ARC Innovation Center, they launched over 20 startups in 2024 alone, with projects being chosen for their potential to improve patient outcomes and advance research, in alignment with the hospital’s goals.
“To make the best out of AI’s potential, you really need the training, the environment and the systems to track and monitor it and educate people on how to use it.” – Dr Gal Goshen
Sheba is now developing an “AI academy” to help shift from simply having AI tools to becoming a hospital built around AI.
Professor Asad Mian, Lead for Healthcare Transformation at Evercare Group (Nigeria and Pakistan), offered a perspective from bringing transformation to lower resource settings, where resource limitations often spark creative, scalable solutions. By using human-centered design and clinical audits, his team co-identifies issues and implements practical, often low-tech changes. In both Evercare’s Lahore and Lekki hospitals, for example, speech-to-text tools and GenAI-enabled triage data visualizations were introduced only after clinicians clearly articulated their needs – tech introduced with a purpose.
“You don’t always need AI, but you do need to understand your pain points. Sometimes, a small process tweak does the job.” – Professor Asad Mian
AI adoption is as much about people as it is about technology. Paula Montgomery, Executive Vice President of Administrative Affairs and General Counsel at Gillette Children’s Specialty Hospital (USA) highlighted the importance of building organizational buy-in and literacy. Starting with foundational education and a clear governance framework, at Gillette they ensure tools are both effective and aligned with internal capacity. Their technology review boards bring together stakeholders from IT, clinical operations, and leadership to evaluate new solutions and provide follow-up monitoring.
“We use early champions and pilot tests to show results and build trust across teams.” – Paula Montgomery
Dmitry Etin, Digital Health Technology Strategist specializing in health data interoperability, reminded participants that governance is not always the enemy of innovation, highlighting how regulation can actually support, not hinder innovation. Standards, such as those under the European Health Data Space, can ensure safety and interoperability – essential for scaling trusted AI tools across systems. He also stressed the importance of collaboration and cooperation across borders and institutions.
“Standardization is what allows systems to speak the same language. It’s what builds trust.” – Dmitry Etin
When asked what’s next, the panel was aligned: in the next 12 months, expect hospitals to capitalize on ‘easy-win’ AI opportunities and to expand use of generative AI in documentation and diagnostics. The panel also shared what they felt hospitals leaders should prioritise to successfully integrate AI tools into daily work: Start with digital literacy. Understand your organization’s needs and you may realize that sometimes a simple solution is better than an advanced one – make sure you are solving problems not creating new ones. Lastly, find some simple use cases that are easily understandable and use these to create momentum.