Why AI Governance Matters: A Conversation with Dr. Sarah Gebauer
Welcome to the first Episode 1 of the Ardexia Podcast where we speak to leaders, clinicians and advocates for better healthcare. This week, we talk to Dr. Sarah Gebauer founder of Validara Health and a physician who uniquely combines clinical practice as an anesthesiologist with AI governance expertise honed at RAND Corporation.
Healthcare organizations are rushing to adopt AI, but most are missing the critical piece that determines success or failure: governance. In our first episode of the Ardexia podcast, I sat down with Dr. Sarah Gebauer, founder of Validara Health and a physician who uniquely combines clinical practice as an anesthesiologist with AI governance expertise honed at RAND Corporation.
The conversation revealed a truth that every healthcare leader needs to understand: the technology isn't the problem. Implementation is.
The National Security Lessons Healthcare Needs
Dr. Gebauer's background at RAND, working on AI governance for national security issues including biosecurity and cyber threats, gives her a perspective most healthcare innovators lack. "A lot of those key learnings are very applicable to medicine," she explained. "How do we know that the tools we're introducing are safe? We always want to maintain our Hippocratic oath and first do no harm, but we also want to encourage innovation."
The fundamental challenge she identified? AI crosses professional boundaries in ways other technologies don't. In healthcare, this manifests as a dangerous disconnect between IT departments and clinical teams. "The CIO level issues within a hospital and when you're implementing AI—we have to do a better job of combining those and making sure those key people talk to each other and understand each other's concerns."
This isn't unique to AI. We've seen it with telemedicine, RPM, and every digital health innovation. But with AI, the stakes are higher and the complexity greater.
The Three Non-Negotiables for AI Implementation
When I asked Dr. Gebauer what healthcare leaders absolutely need before adopting AI, she was clear:
1. People who actually understand AI Not just IT staff or excited executives, but a dedicated team with adequate time, administrative support, and hospital leadership backing to shepherd the process properly.
2. A system for questions and communication Both for the evaluation questionnaires and for managing communication across the inevitable silos that will be involved. Every hospital is currently "growing their own garden" of AI governance, which is inefficient and risky.
3. A feedback loop This is where most implementations fail. You need mechanisms to hear from the people actually using the tools—clinicians and staff—and that information must flow back to the product creators. "If we don't radically change our approach to feedback and including clinicians and frontline workers in how we use and refine AI, we're going to have the same issues we had with EHR implementation all over again."
Why Clinician Trust Is Everything
The conversation turned to the adoption crisis we see across digital health. Dr. Gebauer pointed out something crucial: the most successful AI tools so far have been the ones that remove burdens, not the ones that claim to make clinical decisions better.
Pre-authorization letters generated by AI? Physicians love them because everyone hates writing them, and AI does them well. Ambient scribes? Successful because they focused on making physicians' lives easier with minimal learning curve.
"This is basic stuff," Dr. Gebauer noted, "but for some reason in healthcare we kind of forget that users like things that are easy to use and that will make our lives easier."
The clinical decision support tools? Those face a much harder road. Why? Because there's a fundamental disconnect between the decision makers' perceived need for these tools and the physicians' perceived need. "It's not a great way to market to an entire profession to tell them they're not doing their job well enough," I added. The tools that succeed will serve both the buyers and the actual users.
What Success Looks Like in Five Years
When I asked Dr. Gebauer about her vision for AI governance success, her answer was patient-centered: "We have figured out how to do AI governance in a way that minimizes the time for good tools to get adopted and into the ecosystem and keeps patients safe."
The promise isn't just about making physicians' lives easier—though that matters. It's about creating solutions for patients that were never possible before. "If you want something to get adopted, create something that patients have a huge need for and doctors will use it because they truly do care about their patients."
The Bottom Line for Healthcare Leaders
Dr. Gebauer's closing advice cuts through all the complexity: "You have to involve clinicians in your process of creation and feedback loops. It's very easy to create a product with only engineers, who are wonderful. But if it's not used by clinicians, then all that promise, all those solutions for patients, they're not going to get adopted and everyone is going to lose."
This aligns perfectly with Ardexia's methodology: prioritize people over technology. Map real workflows, not documented ones. Design integrations that feel natural to clinicians. Build feedback loops that actually work.
The healthcare organizations that will successfully adopt AI aren't the ones with the biggest budgets or the flashiest technology. They're the ones that understand governance, involve clinicians from the start, and build for the humans who will use these tools every day.
———————-
Thanks for listening! More information on this episode:
Learn more about Dr. Sarah Gebauer's work and follow her insights : Validara Health | Substack | LinkedIn
Watch the full conversation on Youtube.
Listen on Spotify or Apple Podcast.
Ready to transform your healthcare innovation approach? Contact Ardexia to discuss how we can help you move from pilot to sustainable adoption.
Dr. Aditi Joshi is the CEO of Ardexia and host of the Ardexia podcast. She's an emergency physician who has built 13 digital health programs across three continents and specializes in turning failed digital health implementations into measurable clinical and financial success.