Reflections from the HIMSS AI in Healthcare Forum today in Brooklyn
TODAY at the HIMSS AI in Healthcare Forum , I had the opportunity to sit in on one of the more thought-provoking discussions of the event: The Future of Healthcare with Agentic AI. The conversation, led by executives from NYU Langone, Hyland, Soothien HealthTech, and MATTER, tackled a key emerging question: how do we move from simple bots and rules-based automation into the domain of autonomous agents in healthcare?
The concept of agentic AI—tools that not only think (via LLMs) but act independently—presents a massive opportunity. But as was made clear, healthcare is not a low-risk industry where trial-and-error is tolerated. Governance, liability, vendor transparency, and data maturity all came up as friction points. As one speaker noted, “80% of hospitals have zero data maturity.” Oof.
What stood out to me most:
- Start small. Don’t wait for the tech to be perfect. Begin with low-variance workflows like pre-op documentation, quality improvement outreach, or resident education. These are fertile ground for early agent deployments.
- Own your outcomes. Clinicians are right to feel the “weight” of decisions—and AI agents won’t (and shouldn’t) remove that. These systems must enhance, not displace, clinical judgment.
- Governance is not optional. AI traceability, observability, and guardrails must be baked into any deployment. Think of agents like new team members—you wouldn’t hire without oversight, feedback loops, and continuous training.
- Vendor partnerships matter. This isn’t just about buying a tool. It’s about co-developing a long-term operating model where risk, drift, and change are managed collaboratively.
I especially appreciated the nuanced view on AI liability and the role of physicians. “That weight is good weight,” one panelist said. “If my residents don’t feel it, that’s a red flag.”
On a related note, I’ll be hosting an evening executive dinner reception tonight nearby in my role as founder of the CxO Security Forum. We’re bringing together hospital CISOs, CIOs, and startup leaders to continue this discussion—specifically on how security, AI governance, and digital trust intersect in the healthcare environment. (If you’re in town and belong in the room, shoot me a note.)
Overall, this was a fascinating glimpse into a future that’s already unfolding. Agentic AI isn’t theoretical anymore—it’s operational, and it demands a new kind of leadership.