Health System Tech Leaders Hit the Brakes on AI

Health systems are deliberately pumping the brakes on artificial intelligence, according to Becker's Hospital Review, which asked nine hospital technology leaders how fast they are moving. Their answer: slower than the hype suggests, and on purpose.

The caution has a clear trigger. In a September essay, Anthropic CEO Dario Amodei warned that AI models are gaining capabilities faster than the industry can verify they are safe. For health systems, that gap is not abstract. The same class of technology is being deployed where errors reach patients directly, from clinical documentation to decision support. Hospital leaders say that raises the bar for validation, monitoring and governance before any tool touches care.

In practice, that means more pilots, tighter guardrails, and a preference for narrow, well-tested use cases over sweeping rollouts. The leaders frame slowing down not as resistance to AI but as the responsible way to adopt it. For a sector where mistakes carry clinical consequences, speed is not the goal. Trust is.