AI Prior Authorization Systems at Major Health Insurers Linked to Patient Deaths and Treatment Denials

In 2023, investigations by the Senate Permanent Subcommittee on Investigations revealed that major health insurers including UnitedHealth Group, Cigna, and Humana had deployed AI-powered prior authorization systems that were systematically denying medically necessary treatments. The systems, designed to reduce costs and streamline approval processes, were found to be making life-and-death decisions with minimal human oversight, contributing to documented patient deaths and serious health deteriorations. The investigation, led by Senator Elizabeth Warren and others, documented specific cases where AI systems denied coverage for critical treatments including cancer medications, emergency surgeries, and mental health interventions. Internal documents revealed that these systems were programmed with aggressive denial targets and were overriding physician recommendations at unprecedented rates. UnitedHealth's AI system alone was processing over 90% of prior authorization requests with minimal human review, with denial rates increasing significantly after AI implementation. Documented patient harm included multiple deaths from delayed cancer treatments, emergency room visits that could have been prevented with approved medications, and deterioration of chronic conditions when AI systems denied routine but necessary care. Families of affected patients reported that appeals processes were lengthy and often unsuccessful, with AI systems maintaining denials even when presented with additional medical evidence. The human cost extended beyond individual patients to families facing financial ruin from uncovered medical expenses. The regulatory response was swift and comprehensive. The Centers for Medicare & Medicaid Services (CMS) proposed new rules in late 2023 requiring transparency in AI-driven authorization decisions and mandating human physician review for denials of certain critical treatments. The proposed regulations also included requirements for insurers to disclose their use of AI in prior authorization and to provide detailed justifications for denials. Multiple class action lawsuits were filed against the insurers, alleging violations of ERISA, state insurance laws, and patient protection statutes, with damages potentially reaching hundreds of millions of dollars.