Radiology Jobs Grow 26% Despite AI Taking 75% of FDA-Cleared Medical Devices
Updated
Updated · Ars Technica · Aug 25
Radiology Jobs Grow 26% Despite AI Taking 75% of FDA-Cleared Medical Devices
1 articles · Updated · Ars Technica · Aug 25
Summary
A decade after Geoffrey Hinton predicted radiologists would be replaced within five years, the specialty is still expanding, with practitioner ranks projected to grow 26% or more over the next 30 years.
About three-quarters of the 1,400 AI-enabled medical devices cleared by the FDA by early 2026 target radiology, making the field healthcare's main testing ground for expert decision systems.
Those tools are reshaping work rather than eliminating it—drafting reports, flagging urgent scans and, in some cases, detecting abnormalities that humans may miss or matching trained radiologists' readings.
Accuracy gains matter because diagnostic-image error rates run about 3% to 5%, or roughly 40 million mistakes worldwide each year; one analysis of 43 trials found AI-assisted colonoscopies detected more polyps than conventional exams.
The emerging lesson is that radiology's future hinges less on humans versus machines than on combining AI's precision with physicians' judgment and flexibility for patient care.