Follow the money in American healthcare and you eventually arrive at a strange fact: there is a multi-billion-dollar industry whose entire business model depends on insurers not paying claims promptly. Denial management — the software, services, and outsourcing built to fight, predict, and prevent claim denials — has become one of the fastest-growing segments in health IT. It exists because the friction exists. If payers paid cleanly, the industry would have no reason to be.
This is the perverse economics at the heart of the revenue cycle. Every percentage point of denial rate is revenue for someone.
The Players
The market leaders are names every hospital CFO knows. Waystar, which went public on Nasdaq in June 2024, sells cloud-based revenue-cycle software including denial prevention and recovery tools to providers processing hundreds of millions of claims. R1 RCM — taken private in 2024 in a $9 billion-class deal — runs end-to-end revenue cycle operations for health systems, effectively outsourcing the entire fight. FinThrive, formed from the merger of nThrive and TransUnion's healthcare business, pitches analytics-driven revenue cycle and denials management. Around them orbit Experian Health, Conifer Health Solutions, Ensemble Health Partners, and a wave of AI-native entrants: AKASA, Infinx, and others selling machine-learning models that predict which claims a payer will deny before the claim goes out.
The market numbers are striking. Future Market Insights sizes the revenue-cycle denials intelligence segment alone — the algorithms and analytics that predict, identify, and resolve rejected claims — at $2.4 billion growing toward $7.8 billion, a 12.5% compound annual growth rate. Broader revenue-cycle software was valued around $12.8 billion in 2026. These are not niche figures. This is a major health-IT category built entirely on payer friction.
The Perverse Incentive Structure
Consider what this means. Hospitals spend an estimated $18 billion a year overturning denials, per the American Hospital Association. A meaningful share of that flows to the denial-management industry — software licenses, outsourced billing operations, consulting engagements, AI platforms. The industry's revenue grows when denial rates grow. Its customers' costs grow when denial rates grow. Everyone in the chain is paid to manage the problem; no one in the chain is paid to eliminate it.
This is not a conspiracy. It is an equilibrium. Payers use initial denials to slow payments and protect margins — Kodiak's data shows initial denial rates climbing from 10.2% in 2020 to 11.8% in 2024, with payers ultimately paying roughly 90% of claims. Providers respond by buying denial-management technology, which improves their recovery rates at the margin. Payers respond by shifting denial logic — authorization denials down 7.7% in 2024, medical-necessity and information-request denials up to replace them. Each side invests. The friction persists. The vendors collect from both directions.
The AI chapter is accelerating the cycle. Early adopters of AI-driven denial prediction report roughly 20% declines in rejected submissions — genuine improvement. But payers are deploying their own AI to auto-deny at scale, and several are now facing regulatory scrutiny and lawsuits over algorithmic denial practices. The arms race is getting more expensive for everyone except the arms dealers.
The AI Arms Race
The newest front in the denial war is algorithmic, and both sides are armed. On the provider side, machine-learning models trained on historical claims data now predict — before submission — which claims a given payer is likely to deny and why. The models flag missing documentation, authorization gaps, and coding mismatches in real time, letting billing teams fix claims pre-bill rather than appealing them post-denial. Vendors like AKASA and Infinx have built their entire pitch around this: denial prevention as a software problem, with early adopters reporting meaningful reductions in rejection rates.
On the payer side, the same technology runs in reverse. Algorithmic claim review can deny claims at a volume and speed no human reviewer could match — and several major payers are now defending lawsuits alleging their AI-driven denials systematically rejected medically necessary care. The legal theory is straightforward: an algorithm that denies at scale, with minimal human review, is not utilization management. It is a business model.
Regulators are circling. CMS rulemaking has begun constraining automated prior-authorization denials in Medicare Advantage, and state attorneys general are investigating whether algorithmic denials violate state insurance law. But regulation moves slowly and software deploys quickly. For the denial-management industry, the AI arms race is the best possible news: every new algorithmic denial creates demand for a counter-algorithm, every payer model spawns a provider model, and the vendors sell to both sides of every escalation.
This is the equilibrium in its purest form. The technology gets better. The friction persists. The industry grows.
What Would Actually Fix It
The honest answer is structural, not technological. Standardized prior-authorization rules with real enforcement. Penalties for denial rates above evidence-based thresholds. Transparency requirements that force payers to publish denial rates by category — sunlight that currently does not exist. Some of this is moving: CMS has tightened prior-authorization rules for Medicare Advantage, and state legislatures keep introducing gold-carding laws that exempt reliable providers from authorization requirements.
Until then, the denial-management industry will keep growing, because the problem it manages keeps growing. It is excellent at its job. That is precisely the problem.
"An industry built on friction has no incentive to end the friction. It has every incentive to manage it — profitably, indefinitely."