Healthcare Leaders Reveal the Foundations of Scalable AI
Governance Defines the Future of AI in Healthcare
With AI now firmly embedded into many health systems, the central question has shifted from "if" to "how" to effectively manage these solutions in clinical practice. This change sets the stage for a deeper conversation about what enables responsible and impactful AI integration.
In a recent webinar, LucidHealth’s Chief Information Officer, Tom Hasley, shared that governance is the key to transforming AI from an experiment into an infrastructure. He noted lasting AI success depends on a strong governance foundation.
For organizations that have scaled AI, governance establishes a framework for evaluating, deploying, and monitoring systems. These committees define ownership and criteria, ensuring each model fits clinical workflows.
At LucidHealth, the internal AI committee reviews each algorithm for accuracy, workflow fit, and long-term sustainability. The committee also determines if the vendor meets clinical infrastructure standards. Beyond selecting and validating the algorithm, the team investigates vendors' data sources, model training practices, update processes, and technology safeguards.
“We’re going into vendor conversations with the governance questions hospitals are asking us—where did you train your data, how are you governing it, how often is it updated,” said Hasley.
Building Clinician Trust Through Proven Value
While early AI deployments prioritized algorithm accuracy, leading health systems have moved beyond technical performance and are now grounding their governance frameworks in measurable, operational impact.
Rad AI’s Chief Innovation Officer, Demetrios Giannikopoulos, emphasized during the webinar, “Early on, it was all about algorithmic characteristics. Now it’s started to morph more into a ‘what are the outcomes of artificial intelligence?”
Bridging accuracy and utility, Matt Zawalich, Vice President of Information Technology Services at Yale, explained that the real measure isn’t a model’s accuracy on paper, but whether the technology genuinely helps people do their work and care for patients.
“ I think that is what gets people engaged, gets them willing to use these things over and over again. When they see that it is making a meaningful difference to their patients and a meaningful difference in terms of how they work, when you establish trust around those tools, you're able to make headway,” said Zawalich.
To hear the full conversation, including deeper insights from LucidHealth and Yale New Haven Health on building a successful, scalable AI strategy, watch the complete webinar HERE.
