AI is writing denial letters. Now it's writing the appeals, too
For Neal Shah, the breaking point came during his wife's cancer treatment. Stacks of bills, denials and statements covered their kitchen table. After totaling premiums, deductibles and uncovered care, they faced a harsh conclusion: carrying health insurance had cost more than paying cash might have.
What started as shock turned into research, then action. Shah built CareYaya to connect families with affordable caregiving from pre-health students. But the pattern kept repeating: people were crushed by claim denials and confused by appeals. The result is Counterforce Health - a free platform that drafts customized appeal letters in minutes, so patients can push back without spending days decoding policies and medical literature.
The denial trend insurance teams can't ignore
Denials have become a defining feature of US health care. According to KFF, marketplace plans denied roughly 20% of claims in 2023, yet fewer than 1% were appealed. That's an imbalance that leaves care on the table and fuels distrust across the system.
Medicare Advantage audits have also flagged widespread, wrongful delays or denials for care requests, affecting tens of millions of beneficiaries. As one clinician put it, some patients step into the ring ready to fight; many more give up, go into debt or skip treatment entirely.
Source: KFF on ACA denials and appeals
AI is already in the loop - on both sides
Major insurers now use AI to scan records, apply rules and generate denials at scale. Investigations have reported near-instant determinations that sometimes conflict with treating clinicians. In late 2023, a class-action lawsuit alleged that an AI model used by a UnitedHealth subsidiary cut off care prematurely for older patients, regardless of doctor recommendations.
Coverage: Reuters on the UnitedHealth algorithm lawsuit
Counterforce flips the script: if AI can automate denials, AI can also automate evidence-based appeals. The goal isn't noise - it's precision. Strong citations, policy references and clinical rationale, assembled fast.
How Counterforce Health works
Upload a denial letter and relevant records. The platform analyzes the insurer's rationale, policy language, and clinical evidence. It also looks at patterns from successful appeals with similar fact patterns.
The output is a complete, editable appeal that addresses medical necessity, coverage criteria and coding details. What used to take eight hours of research and drafting takes minutes. Users can revise and submit through the insurer's usual process.
Why access and cost matter
Most people don't appeal - not because they're wrong, but because the process feels impossible. Shah puts it bluntly: "Ninety-nine percent of patients or family caregivers do not appeal. Of the tiny fraction who do, 40% win." That's a signal that friction - not evidence - decides outcomes far too often.
Counterforce is free for individuals, backed by grants and venture support, including a $2.47 million award from PennAITech at the University of Pennsylvania. Patient advocates can help, but at $80 to $150 per hour, that option is out of reach for many households already stretched by medical bills.
Implications for insurance professionals
AI-generated appeals are going to become standard. That changes the rhythm of UM, SIU and appeals units. You'll see a higher volume of higher-quality appeals that directly quote your policies, cite clinical criteria and challenge inconsistent determinations.
This isn't a threat; it's an opportunity to raise the floor on accuracy and trust. If patients can generate clear, evidence-driven appeals, plans can respond with clearer determinations, tighter policy language and better member communication. Over time, that reduces abrasion and rework.
What to do next
- Audit the top denial codes and reasons that are frequently overturned. Tighten criteria and education where overturn rates are high.
- Standardize policy language so it's explicit and machine-readable. Expect appeals to quote it verbatim.
- Publish member-facing guidance on how to appeal effectively. Transparency reduces friction and repeated submissions.
- Instrument your appeals funnel. Track time-to-determination, overturn rates by reason code and member sentiment post-appeal.
- Align medical policy, UM and compliance on AI use. If models inform denials, document oversight, sample reviews and exception handling.
- Pilot clinician-to-clinician peer review windows that preempt avoidable denials before they harden into appeals.
A clinician's view from the other side of the table
Dr. David Casarett, a palliative care leader at Duke, saw his own mother face repeated denials for necessary cancer medications. His family had the expertise and time to contest them - and they often won. His question lingers: what chance does a single parent working two jobs have in the same fight?
That's the practical gap tools like Counterforce aim to close: less bureaucracy, more clinical focus, and quicker access to care that meets policy and medical necessity standards.
Privacy and security
Appeals require sensitive data. Counterforce states it uses encryption, does not sell personal information and shares data only with trusted providers under confidentiality and legal requirements. As with any service, teams should review the privacy policy, data retention practices and BAAs where appropriate.
How to get started with Counterforce Health
Visit the site and select "Start Free Appeal." Complete a short questionnaire about your denial and upload the denial letter plus insurance details.
The system generates a full draft appeal. You can edit and submit through your plan's existing channels. For clinics, this can slot into front-desk or revenue cycle workflows to reduce write-offs and delays.
Leveling up AI literacy on your team
If you're building internal skills around AI policy, claims operations or compliance, you may find curated programs useful. Explore role-based learning paths here: AI courses by job.
Your membership also unlocks:
