The Trump administration is using AI to analyze Medicare prior authorizations, and companies get 25 percent of the “savings” the system generates.
We found two private equity firms, Carlyle Group and Patient Square Capital, backing this process.
Video by Eric Gardner and Eric Schuman
TRANSCRIPT
ERIC GARDNER, MORE PERFECT UNION: I’ve been trying to figure out: What is AI actually doing to American healthcare? I didn’t expect to find the answer on a utility bill from India.
I’ve always been very interested in how AI is reshaping American healthcare, and I found an answer in an unlikely place. A utility bill. From India.
It’s for an address here, which connects back to a company hired by the federal government to harness ‘artificial intelligence and machine learning’ to improve American healthcare.
ELON MUSK CNN CLIP: When you have superintelligence, it means that everyone in the world can have better medical care than anyone here, including me.
GARDNER: That’s what we’re being promised. But how is AI being used in healthcare today?
In January, the government launched a program that uses artificial intelligence to help make decisions for hundreds of thousands of patients.
CAREL EDGERLY, MICHAEL EDGERLY’S WIFE: The AI apparently did not read the chart notes.
MICHAEL EDGERLY, CAREL’S HUSBAND: Chances are maybe I could be, I could be walking instead of using a walker.
GARDNER: When I started to investigate, I realized how little the public actually knew about the program.
ZENA WOLF, SENIOR ADVISOR AT THE CENTER FOR HEALTH AND DEMOCRACY: There is so little information about the companies, the contracts [Centers for Medicare and Medicaid] has, and the algorithms that they’re using.
GARDNER: And when I tried to find out who built the system making some of these decisions, I discovered a maze of companies, some that barely exist, and investors behind it making money no matter the patient outcomes.
The whole thing led me here.
A very strange place to ask the question: What is AI actually doing to American healthcare?







