Americans don't trust AI, and they trust health insurers even less.
UnitedHealth is betting $3 billion that one can fix the other.
The country's largest health insurer says its AI push is already returning two dollars for every one it spends.
We track the moves Wall Street is actually paying attention to in Market Briefs - five minutes every morning, plus a free investing masterclass when you join.
UnitedHealth plans to spend $3 billion on AI in 2026 and 2027, and the early math is what Wall Street wants to see. Leaders say they're getting a 2-to-1 return as AI takes over slow manual work and frees up workers.
The company also expects to cut nearly $1 billion in costs this year, mostly from AI.
The stock has noticed: shares are up 21% this year after falling more than a third in 2025, when profits collapsed and UnitedHealth became the face of nationwide anger at the insurance system.
Most industries went digital years ago, but healthcare didn't.
Insurers and providers still spend $80 billion a year on administrative transactions - phone calls, faxes, and forms moving between offices, per a Morgan Stanley report.
UnitedHealth's Optum Real system lets doctors check in real time whether a service is covered, and it's already run about a billion checks since launching last year.
The company also has AI agents calling doctors' offices to book patient appointments, AI reading chart summaries to nurses on the way to home visits, and software scanning millions of customer calls to spot what people complain about most.
Behind all that, UnitedHealth runs more than a thousand AI uses, employs 20,000 AI engineers, and has 117 large language models its workers can pull from.
The harder part is the public.
A Gallup survey last year found 69% of Americans have little or no trust in businesses to use AI responsibly. Health insurers start with an even steeper hill, with about half of insured Americans saying they've hit trouble getting care like delays and denials, per KFF.
UnitedHealth is already facing class action lawsuits claiming it and others used an algorithm from its naviHealth subsidiary to limit care, including admissions to post-acute care after hospital stays. A federal inspector general report found companies using the algorithm had higher rates of denials - and that those denials were almost always overturned when patients appealed.
The company disputes the claims and says the algorithm wasn't used to make coverage decisions.
The backlash got more personal after the 2024 killing of UnitedHealth insurance chief Brian Thompson. Since then, the company and its rivals have been rolling back some prior approval requirements under pressure from Washington.
UnitedHealth is being clear: almost 99% of its AI work is administrative, not clinical. The company says it isn't building diagnostic AI, and an internal review board of medical ethicists, clinicians, technologists, and privacy and legal experts signs off on new uses.
The 2-to-1 return is the easy part.
If you want this kind of read on the market every weekday, join 350,000+ investors reading Market Briefs - you also get a 45-minute investing course thrown in as a bonus.