Nebraska Medicine Kills AI Cancer Registry Pilot

Nebraska Medicine set out to automate one of its most labor-intensive tasks: pulling data from patient charts to feed its cancer registry. According to Becker's Hospital Review, the Omaha-based health system pulled the plug on the AI pilot just two months in.

The math explains why the system tried. The registry takes in roughly 6,000 new cases each year, and each case takes a person about an hour of chart abstraction. That is thousands of nursing hours spent on tedious, high-stakes data work. But cancer registry data must be precise, since it feeds reporting, research, and quality tracking, and the tool could not clear that accuracy bar.

The decision reflects a broader pattern. Hospitals are racing to apply AI to administrative and documentation burdens, but many pilots stall when generative tools cannot match human reliability on structured, consequential data. Nebraska Medicine's willingness to publicly kill a pilot quickly, rather than force adoption, is itself a signal of disciplined AI governance that more systems are adopting.