Court Allows Discovery Into Insurer’s Use of AI to Deny Claims

Published
Score
8

Why it matters

A Minnesota federal court ruled on March 9, 2026, in the Lokken case, granting plaintiffs' motion to compel discovery from UnitedHealth Group (UHC) into its use of the AI tool nH Predict—developed by Optum subsidiary naviHealth—for evaluating and denying Medicare Advantage post-acute care claims.[1][2][4] The court approved broad production of documents on nH Predict's development, goals, use policies, employee training, oversight, and certain government investigations, but denied requests for source code, internal probes, financial data, and employee incentives.[1][2] UHC maintains nH Predict is a non-generative care-support tool aiding recovery planning, with final decisions by physicians per CMS guidelines.[2]

Plaintiffs, families of two deceased Medicare Advantage members, filed the class-action lawsuit in 2023, alleging UHC systematically denied coverage using AI to override clinical judgment and create financial gains, harming patients.[1][2][3][4] Involved parties include UHC, Optum/naviHealth, and U.S. Magistrate Judge in the District of Minnesota; the ruling applies principles beyond health insurance to general coverage disputes.[1]

This stems from rising insurer AI adoption for claims processing, amid lawsuits claiming wrongful denials (e.g., 80% of Medicare Advantage denials later approved due to documentation issues).[1][7] Timeline: suit filed 2023; motion compelled March 2026; reported March 10-23, 2026.[1][2][4] Newsworthy now due to escalating AI battles—insurers/hospitals spent $1.4B on AI in 2025, UHC planning $1.5B in 2026 for savings—sparking legislative pushes like Minnesota HF2500 to ban AI in prior authorization denials from 2027, plus broader litigation trends.[3][5][7]

Sources

mail

Get notified about new Artificial Intelligence developments

Primary sources. No fluff. Straight to your inbox.

See more entries tagged Artificial Intelligence.