RadNet’s AI bet shows radiology is moving from pilots to platform economics

RadNet’s aggressive AI strategy is notable because it comes from an operator, not just a vendor. That matters: providers often understand where the inefficiencies live, and they are in a better position to measure whether AI actually saves time, reduces cost, or improves patient flow.

The company’s bet suggests radiology is entering a platform phase. Instead of treating AI as a series of isolated add-ons, major imaging groups are trying to weave it into the economics of the business — how studies are scheduled, read, escalated, and billed.

But this strategy also raises the bar for proof. A workflow tool may be technically impressive and still fail if it complicates the reader experience, creates alert fatigue, or does not integrate with real-world throughput constraints. Operational AI only works if clinicians trust it enough to let it disappear into the background.

If RadNet succeeds, it could provide a template for other large specialty providers: own the workflow layer, standardize the data, and use AI to turn scale into an advantage. If it falls short, it will reinforce a familiar lesson in healthcare AI — that deployment complexity often matters more than model performance.