DocumentedAlgorithmic Care Denial (Healthcare)November 2023United States

UnitedHealth nH Predict AI Allegedly Denied Care to Elderly Patients

A class action (Estate of Lokken v. UnitedHealth Group) alleges UnitedHealthcare used an AI model called nH Predict to cut off post-acute care for elderly Medicare Advantage patients, overriding treating physicians. Plaintiffs allege the model's predictions were reversed roughly 90 percent of the time on appeal, and that the company relied on the low rate at which patients appeal (about 0.2 percent).

AI system:nH Predict (developed by naviHealth)

Impact

A leading US test case for AI accountability in healthcare. A STAT News investigation first surfaced the practice; other plans reportedly used the same tool.

Outcome

A judge dismissed several counts but allowed the case to proceed. On March 9, 2026 a federal magistrate judge compelled UnitedHealth to disclose internal documents on the algorithm. UnitedHealth denies wrongdoing and says decisions follow Medicare criteria.

Sources

  1. Source 1CBS Newswww.cbsnews.com/news/unitedhealth-lawsuit-ai-deny-claims-medicare-advantage-health-insurance-denials/
  2. Source 2The Registerwww.theregister.com/2023/11/15/unitedhealthcare_ai_medicine/
  3. Source 3Healthcare Finance Newswww.healthcarefinancenews.com/news/unitedhealth-ai-algorithm-allegedly-led-medicare-advantage-denials-lawsuit-claims

Related incidents

Same category, country, or harm tier.

United States·July 2023
Cigna PxDx Algorithm Allegedly Used to Deny Claims in Bulk
Following a ProPublica investigation, a class action (Kisting-Leung v. Cigna) alleged Cigna used an automated system called PxDx to reject large batches of claims, with company doctors signing off without opening patient files. ProPublica reported more than 300,000 payment requests denied over two months in 2022, averaging about 1.2 seconds per claim. CAVEAT: Cigna publicly disputes that PxDx involves AI or machine learning, describing it as long-standing code-matching technology. This entry is retained with that dispute stated explicitly; it sits at the boundary of AI versus automated algorithm.
United States·May 5, 2026
Pennsylvania Sues Character.AI Over Chatbots Posing as Licensed Doctors
Pennsylvania Governor Josh Shapiro's administration sued Character.AI, alleging unlawful practice of medicine. A state investigator posing as a patient interacted with a persona named 'Emilie' that claimed to be a psychiatrist, said it attended Imperial College London, claimed to be licensed in Pennsylvania, and supplied a license number the state confirmed was invalid.
United States·April 2026
AI Voice-Clone Cartel Virtual Kidnapping of Las Vegas Mother
In April 2026, a Las Vegas mother received what she believed was a kidnapping call from her daughter, sobbing on the line. A male caller then claimed cartel custody and demanded $15,000. The 'daughter's voice' was an AI clone built from publicly available social-media audio. The caller kept the mother on the phone for nearly six hours while she drove between banks, grocery stores, and a Walmart, wiring money to Toluca, Mexico.
← Back to all incidentsSee full catalogue