Lawmakers and Advocates Press White House to Halt AI Insurance Denials

The investigation highlights how private insurers, which manage federally funded Medicare Advantage plans, increasingly rely on predictive software to set hard limits on patient care. These systems determine when a patient should leave a hospital or nursing home, often ignoring the reality of the individual’s recovery. According to reports, these denials are frequently triggered without human oversight, leaving patients—many of whom are suffering from cancer or recovering from surgeries—to either fund their own care or face a grueling, months-long appeals process.

Industry giants including UnitedHealth Group, Elevance, Cigna, and CVS Health have spent recent years acquiring the companies that develop these predictive tools. One such model, NaviHealth, gained notoriety after a UnitedHealthcare algorithm used its data to cut off payment for 89-year-old Dolores Millam just 15 days after her surgery. Despite her inability to move or perform basic tasks, her coverage was revoked, forcing her family into a federal court battle that eventually overturned the denial—but only after $40,000 in bills had accumulated.

Critics argue that these practices prioritize corporate profit over patient health. Senator Warren has pointed to the dual abuse of "upcoding," where insurers inflate patient illness to secure higher government payments, while simultaneously using AI to restrict actual treatment. With seven companies now controlling over 70% of the Medicare Advantage market, advocates are meeting with White House staff this week, insisting that the Biden administration must establish strict regulatory guardrails to end what they describe as a barbaric reliance on automated, profit-driven care rationing.