THE DENIAL
MACHINE.

What happens in the 1.2 seconds between claim submission and AI denial. The float mechanics. The investment return accruing while the denial is issued. The documented architecture of the insurance premium.

Cigna PxDx. UnitedHealth nH Predict. The documented AI systems processing and denying claims at scale. How they were built, what they optimize for, and what the patent filings say about their intended function.

Approximately 1% of denied claims are ever appealed. Despite the fact that the reversal rate — when appeals are filed — is high. The machine is optimized for volume and attrition, not accuracy. This chapter documents why.

The documented history of how the American health insurance system was structured — and who structured it. The foundation that built the system still operating today. The receipts go back further than most people expect.

Who owns the insurers. Who owns their investors. How the same institutional capital that appears across every other Base Layer HQ investigation appears here — in the companies processing and denying your health insurance claims. The ownership chain from your premium to Wall Street — assembled for the first time in one place. All public record.

Base Layer HQ follows the documents wherever they go. This chapter documents the official case for AI in insurance claims processing, the documented error rates from active litigation, and what the industry's own regulatory filings reveal about the gap between the public argument and the operating reality.

The Official Argument. The health insurance industry argues AI delivers three benefits: efficiency, fraud detection, and cost reduction. A 2024 NAIC survey found 84% of large health insurers use AI operationally. A 2024 J.D. Power survey found digital claims satisfaction up 17 points year-over-year. Stanford Health AI researchers documented in January 2026 that AI could reduce denials caused by incomplete administrative submissions — a real problem that affects patients and providers. The industry's public position is that AI removes human error and bias from claims review, and that faster processing benefits patients.

The nH Predict Class Action. A class-action lawsuit filed in the US District Court for Minnesota alleges that UnitedHealth and naviHealth used the nH Predict algorithm to systematically deny rehabilitative care for Medicare Advantage patients — with error rates as high as 90%. Most patients who appealed their denials won reversals. But only approximately 1% of denied patients ever appeal. In February 2025, a federal court allowed the class action to proceed on breach of contract and good faith claims. Court documents are on PACER.

The Appeal Rate Document. Health Affairs (January 2026) documented that approximately 1% of denied claims are appealed — despite high reversal rates when appeals are filed. Insurance companies design denial systems knowing that number. The algorithm's value is not accuracy. It is volume and attrition. A 90% error rate is not a malfunction when 99% of people will not challenge the result.

The Regulatory Response. CMS issued a rule in January 2024 requiring Medicare Advantage plans to base coverage decisions on individual circumstances — not algorithmic predictions — with physician review required. Issued specifically in response to documented AI denial practices. The insurance industry opposed these safeguards. Their documented lobbying position — that physician review creates administrative burden — is on file with CMS. Both the position and the final rule are public record.

The Reform Side's Own Documented Disagreement. Stanford researchers documented in 2026 that AI could reduce administrative denials — claims rejected due to incomplete paperwork rather than coverage decisions — which represent a significant share of total denials. The argument is that well-designed AI with proper human oversight could produce better outcomes than the current prior authorization system. CMS's 2024 rule was designed to preserve AI efficiency while mandating human accountability. The receipts exist on both sides.

The following strategies exist in documented public record. This chapter documents their existence and access mechanics. It is not legal or financial advice. Verify independently before acting.

1. Appeal Every Denial. The documented reversal rate when appeals are filed is high — and the industry depends on 99% of people not knowing that. Under your plan's Summary of Benefits and Coverage — a document insurers are legally required to provide — the internal appeal process is described. File the internal appeal first, in writing, citing the specific coverage provision. If the internal appeal fails, you have the right to an external independent review under the ACA. The external review process is documented at healthcare.gov/marketplace-appeals.

2. Request the Clinical Criteria. Under ERISA and state insurance laws, you have the right to request the specific clinical criteria used to deny your claim. This is the algorithm's decision logic — the documented standard your claim was measured against. Insurers are required to provide it. Request it in writing. If the criteria applied don't match your plan's coverage documents, that discrepancy is the basis of your appeal. This strategy is documented in the Patient Advocate Foundation's claims resource library at patientadvocate.org.

3. File a State Insurance Commissioner Complaint. Every state has an Insurance Commissioner with jurisdiction over claims practices. A documented pattern of AI denials without physician review is a claims handling violation in most states. Filing a complaint creates a public record, triggers a regulatory response, and — critically — insurers are required to respond. The National Association of Insurance Commissioners maintains a state-by-state directory at naic.org. This is the same regulatory body whose 2024 survey documented 84% AI adoption.

4. The No Surprises Act. Effective January 2022, the No Surprises Act prohibits surprise billing for emergency services and certain out-of-network care. If you received a bill that may be covered under this act, the federal dispute resolution process is at federalnosuprises.hhs.gov. The act also requires insurers to provide price estimates for scheduled services. These rights exist in federal law and are enforceable.

5. Know Your Employer Plan Rights. If your insurance is through your employer, the plan is governed by ERISA — the Employee Retirement Income Security Act. ERISA plans are required to provide a full and fair review of denied claims. The Department of Labor enforces ERISA plan compliance. If your employer's plan is self-insured — meaning the employer pays claims directly and just uses the insurer for administration — the denial appeals process follows different rules. Your Summary Plan Description will say which applies. DOL enforcement resources are at dol.gov/agencies/ebsa.