Skip to content

AI Outlooks

News and viewpoints on the latest in AI security

Primary Menu
  • Home
  • What’s new in AI
    • AI Security News
    • Agentic AI News
    • AI Regulation News
    • AI Research News
    • AI Model News
  • Solutions
  • Cybersecurity
    • AI security
    • OWASP
    • Ransomware
    • Shadow AI
  • Learn
    • AI security
    • LLM security
    • AI governance
    • AI compliance
    • Agentic AI
    • AI infrastructure
    • AI data security
  • Home
  • News
  • Medicare paying AI companies to deny claims: New payment model explained
  • Healthcare
  • Ethics
  • Government
  • News

Medicare paying AI companies to deny claims: New payment model explained

Medicare pilots paying AI companies based on rejected claims.
Staff September 8, 2025
health

Medicare will pilot an AI-powered prior authorization system that pays companies based on how many claims they reject, marking a troubling shift toward automated healthcare denials that already plague private insurers.

What is Medicare’s WISeR model?

Medicare’s new Wasteful and Inappropriate Service Reduction (WISeR) model reads like something from dystopian fiction. Starting January, AI companies will earn a cut of the money they save by blocking Medicare beneficiaries from their healthcare.

The program essentially creates what one advocate calls “bounty hunters” in the healthcare system.

The current state of AI healthcare denials

The timing couldn’t be worse. Private insurers already deny claims at record rates, with initial denial rates jumping from 10.15% in 2020 to nearly 12% by late 2023. Medicare Advantage plans show even grimmer statistics, with denial rates rising 56% between 2022 and 2023.

These aren’t just numbers on spreadsheets. They represent cancer patients waiting months for treatment while their conditions worsen, elderly patients kicked out of rehabilitation facilities, and families depleting life savings.

How private insurers use AI to reject claims

The private insurance playbook offers a preview of what’s coming. Cigna’s PXDX system lets doctors reject 60,000 claims monthly without opening patient files, spending roughly 1.2 seconds per case. UnitedHealthcare allegedly uses an AI tool with a known 90% error rate.

When appeals succeed 80-90% of the time, you’re not looking at quality control. You’re looking at systematic obstruction betting that sick people won’t fight back.

Here’s the cruel genius: insurers know exactly what they’re doing. Internal documents show companies predict billions in savings from these AI systems. They count on patient exhaustion. Only 0.2% of denied claims get appealed.

The math is coldly profitable.

Problems with human oversight in AI denials

Medicare officials promise human clinicians will make final denial decisions, but that safeguard rings hollow.

Cigna doctors already rubber-stamp AI recommendations without review. When your performance metrics reward speed over accuracy, when you’re processing 50 denials in 10 seconds, human oversight becomes theater.

Technical limitations and bias in medical AI

The technology itself remains deeply flawed. Recent research shows AI medical systems can go haywire from a single typo, exhibit gender bias by telling women not to seek care more often than men, and fundamentally lack the nuance for complex medical decisions.

Yet we’re handing these systems unprecedented power over healthcare access while their owners profit from every rejection.

Why Medicare’s approach contradicts its mission

What makes this especially galling is the stated justification. Yes, healthcare waste exists. But turning AI companies into commissioned denial agents isn’t reform, it’s weaponizing technology against patients.

California recently passed legislation requiring human physicians to review all AI-driven denials, recognizing the danger. Meanwhile, Medicare marches in the opposite direction.

Geographic scope and expansion risks

The six pilot states – New Jersey, Ohio, Oklahoma, Texas, Arizona, and Washington – become testing grounds for a model that could expand nationwide.

Officials claim they’ll only target “wasteful” procedures like steroid injections for pain relief. But once the infrastructure exists, once companies taste those billions in potential savings, containment becomes fantasy.

This represents a fundamental betrayal of Medicare’s promise. The program exists precisely because private insurance failed vulnerable populations. Now Medicare imports those same failures, dressed up as innovation.

The elderly and disabled, who fled to Medicare for refuge from private insurance cruelty, find the walls closing in.

Future implications for Medicare beneficiaries

Medicare’s AI experiment signals a dark turn for American healthcare. When government programs start paying bounties for healthcare denials, we’ve crossed a line.

The question isn’t whether this technology will expand beyond its initial scope – history shows it will. The question is whether we’ll accept a future where algorithms and profit incentives determine who gets care and who gets abandoned.

The infrastructure being built today won’t stay confined to six states or twelve procedures. Once normalized, automated denial becomes the default. Medicare beneficiaries deserve better than becoming test subjects for Silicon Valley’s healthcare disruption.


FAQs

What is Medicare’s WISeR model and how does it work?

Medicare’s Wasteful and Inappropriate Service Reduction (WISeR) model launches in January, paying AI companies a percentage of money saved by blocking Medicare beneficiaries’ healthcare claims. Companies essentially become “bounty hunters” earning profits from healthcare denials.

How are private insurers currently using AI to deny claims?

Private insurers use AI systems like Cigna’s PXDX, which allows doctors to reject 60,000 claims monthly in 1.2 seconds per case without reviewing patient files. UnitedHealthcare allegedly uses AI tools with 90% error rates to systematically obstruct claims.

What are the main problems with AI-driven healthcare denials?

AI medical systems exhibit severe flaws including going haywire from single typos, showing gender bias against women seeking care, and lacking nuance for complex medical decisions. Human oversight often becomes meaningless theater when doctors rubber-stamp AI recommendations for speed.

Which states will pilot Medicare’s AI denial program?

Six states will serve as testing grounds for the WISeR model: New Jersey, Ohio, Oklahoma, Texas, Arizona, and Washington. Officials claim they’ll only target “wasteful” procedures like steroid injections for pain relief initially.

Why does this program contradict Medicare’s original mission?

Medicare was created because private insurance failed vulnerable populations, yet now imports those same failures. The program transforms from protecting elderly and disabled patients into weaponizing technology against them through commissioned denial agents seeking profit from healthcare rejections.


Tags: Medicare

Continue Reading

Previous: What is the difference between NLP vs LLMs?
Next: How to stop wasting tokens in Claude Code: 7 data-backed fixes

More in AI security

  • Guide

The agentic AI security checklist: 12 controls to verify before you deploy

Staff September 4, 2026
Twelve controls to verify before you deploy an AI agent, each mapped to an OWASP ASI risk...
Read more Read more about The agentic AI security checklist: 12 controls to verify before you deploy
LLM jailbreak defense: techniques that actually stop attacks Jailbreak defense
  • Cybersecurity

LLM jailbreak defense: techniques that actually stop attacks

Staff July 28, 2026
How do enterprises secure AI data pipelines at production scale? safety
  • Cybersecurity

How do enterprises secure AI data pipelines at production scale?

Staff July 28, 2026
How companies can defend against AI model extraction attacks
  • Guide

How companies can defend against AI model extraction attacks

Staff July 23, 2026
What is a model inversion attack?
  • Glossary

What is a model inversion attack?

Staff July 22, 2026

Glossary

model router
  • LLMs

What is a model router for AI? A plain-English guide

Staff July 30, 2026
A model router for AI is a decision layer that picks which large language model answers each...
Read more Read more about What is a model router for AI? A plain-English guide
What is agentic SDLC?
  • Glossary

What is agentic SDLC?

Staff July 22, 2026
What is a model inversion attack?
  • Glossary

What is a model inversion attack?

Staff July 22, 2026
LLM system prompt leakage: what it is, how it works, and how to stop it agentic ai
  • Glossary

LLM system prompt leakage: what it is, how it works, and how to stop it

Staff July 15, 2026
What is LLM supply chain security? (OWASP LLM03:2025 explained) llm supply chain
  • Glossary

What is LLM supply chain security? (OWASP LLM03:2025 explained)

Staff July 14, 2026

Guides

The agentic AI security checklist: 12 controls to verify before you deploy
  • Guide

The agentic AI security checklist: 12 controls to verify before you deploy

Staff September 4, 2026
LLM jailbreak defense: techniques that actually stop attacks Jailbreak defense
  • Cybersecurity

LLM jailbreak defense: techniques that actually stop attacks

Staff July 28, 2026
How do enterprises secure AI data pipelines at production scale? safety
  • Cybersecurity

How do enterprises secure AI data pipelines at production scale?

Staff July 28, 2026
How companies can defend against AI model extraction attacks
  • Guide

How companies can defend against AI model extraction attacks

Staff July 23, 2026
What is a model inversion attack?
  • Glossary

What is a model inversion attack?

Staff July 22, 2026
How to prevent adversarial attacks on AI models
  • Guide

How to prevent adversarial attacks on AI models

Staff July 22, 2026
  • Home
  • What’s new in AI
  • Solutions
  • Cybersecurity
  • Learn
Copyright © All rights reserved. | by AF themes.