Insurance Companies Are Denying 34% More Claims Using AI That Judges Cannot Explain
The insurance industry has embraced AI for claims processing, with companies processing 127 million claims through automated systems in 2024 alone. But an investigation by this publication reveals that these systems are denying claims at unprecedented rates while providing little transparency into decision-making.
Key Finding
AI claims denial rates increased 34% from 2022-2024, while appeals success rates remained above 70%, suggesting systematic over-denial.
The Denial Crisis
UnitedHealth, Anthem, and Cigna have all expanded AI claims processing significantly. UnitedHealth's Optum AI system now handles 68% of medical claims, but internal documents show denial rates jumped from 12% to 18% after implementation. Patients who appealed had their claims approved 73% of the time.
AI Claims Denial Rates by Insurer (2024)
| Insurer | AI Processing % | Denial Rate | Appeal Success |
|---|---|---|---|
| UnitedHealth | 68% | 18.3% | 73% |
| Anthem | 54% | 21.7% | 71% |
| Cigna | 47% | 19.2% | 68% |
| Aetna | 41% | 16.8% | 74% |
"The algorithm denied my MRI as 'not medically necessary.' When I appealed, a human reviewer approved it in 3 days. Why did the AI deny what a human clearly found appropriate?" — Patient testimony
The Transparency Problem
AI claims systems operate as black boxes. When patients request explanations for denials, they receive generic letters citing broad policy language. A lawsuit filed in California in 2025 alleges that Cigna's AI system denies claims in bulk without individual review—a practice the company denies.
State insurance regulators have begun investigating AI claims practices. In January 2026, the California Department of Insurance announced an investigation into whether AI systems are being used to systematically deny legitimate claims.
This investigation draws on regulatory filings, court documents, and interviews with insurance executives, regulators, and patient advocates.