With payer underpayments draining over $130 billion annually from U.S. healthcare provider organizations, AI is emerging as the critical technology for identifying hidden revenue leakage-and human expertise remains essential for getting it back.
The Underpayment Crisis AI Was Built to Solve
A healthcare underpayment occurs when a payer reimburses a provider less than the contractually agreed-upon rate for a delivered service. Unlike a claim denial – where the payer rejects a claim outright – an underpayment is a partial payment that falls short of what the provider is owed. Because underpayments arrive as deposits rather than rejections, they frequently go undetected, making them one of the most insidious and costly forms of revenue leakage in the healthcare revenue cycle.
The scale is enormous. According to the American Hospital Association’s 2025 “Cost of Caring” report, Medicare and Medicaid underpaid U.S. hospitals by approximately $130 billion in 2023. Medicare alone reimbursed just 83 cents for every dollar hospitals spent on patient care, and two-thirds of hospitals now operate with negative Medicare margins. Cumulative government underpayments over the second half of the last decade exceeded half a trillion dollars.
Commercial payers compound the problem. The American Medical Association reports a 19.3% claims-processing error rate among commercial health insurers, and industry estimates from MGMA and HFMA suggest providers lose 1% to 11% of net patient revenue annually to underpayments. Internal 2026 data from MD Clarity on management services organizations (MSOs), show 3-5% of net revenue is lost to underpayments.
A 2025 analysis of 117 providers found that more than 32% of medical claims were underpaid, representing over $5 billion in uncollected reimbursements across that sample alone.
Medicare Advantage plans have become an especially acute source of underpayment. A 2025 Health Affairs study found MA plans have adopted policies that approve inpatient admissions but reimburse at observation-level rates – effectively burying underpayments in remittance advices where they are classified as routine contractual adjustments, leaving providers with no formal appeal pathway. At least 40 health systems dropped or declined to renew MA contracts in 2025, with institutions including Mayo Clinic and Johns Hopkins Medicine citing inadequate reimbursement as a driving factor.
Why Traditional Methods Fail – and Why Technology Changes the Equation
For decades, underpayment detection relied on manual audits, spreadsheet-based contract tracking, and rules-based billing systems that could only flag the most obvious discrepancies. These approaches fail at scale for several reasons.
A mid-size health system may process millions of claims annually, each governed by a different combination of payer contract terms, fee schedules, carve-outs, escalator clauses, and mid-level modifiers. Manually auditing even a fraction of these payments against contractual terms is operationally infeasible. Many underpayments hide in zero-balance accounts that appear fully resolved, or are misclassified as contractual adjustments rather than flagged as variances. Without digitized contract modeling, systematic underpayment patterns across thousands of claims from a single payer can go unnoticed for years.
Automated underpayment detection fundamentally changes this dynamic. Modern platforms ingest and digitize all payer contract terms – including complex multi-procedure logic, modifier-based reimbursement rules, and escalator schedules – then automatically compare every incoming payment to expected reimbursement at the line-item level. Intelligent auditing systems can identify patterns that no human auditor could detect across millions of claims: a payer consistently underpaying a specific CPT code by 3% across all locations, a fee schedule that was never updated after a contract amendment, or a modifier that triggers incorrect bundling logic.
The result is a shift from reactive, sample-based auditing to continuous, comprehensive payment validation – turning underpayment detection from a periodic project into an always-on financial safeguard that can lead to substantial revenue recovery. Using MD Clarity’s RevFind platform, an orthopedics management services organization with over 200 physicians across several states uncovered $10.3 million in underpayments from just seven payers.
AI Alone Is Not Enough: The Case for Combining Technology with Human-Powered Recovery
While AI excels at identifying underpayments at scale-surfacing variances that manual processes would never catch-recovering that revenue requires a different skill set. Payer appeals involve navigating more complex contractual language, meeting notification deadlines that vary by plan, composing clinically supported arguments, and escalating disputes through the appropriate channels. This is where experienced, human-powered revenue recovery services become indispensable.
The most effective approach combines tech-driven detection with dedicated recovery specialists who understand payer behavior, reimbursement policy, and the nuances of contract enforcement. Technology surfaces the opportunities and prioritizes them; human experts execute the recovery, negotiate with payers, and ensure that systemic issues feed back into contract renegotiation strategy.
This hybrid model-automation for detection and prioritization, human expertise for recovery and negotiation-creates a compounding effect. Technology ensures nothing is missed; people ensure that what is found gets collected. Organizations that deploy both capabilities simultaneously recover significantly more revenue than those relying on either approach in isolation.
The Path Forward for Healthcare Revenue Cycle Leaders
Underpayments are no longer a back-office nuisance – they are a strategic financial issue that directly impacts operating margins, capital availability, and a provider organization’s ability to invest in patient care. With government underpayments exceeding $130 billion, commercial error rates approaching 20%, and MA plans continuing to adopt policies that suppress reimbursement, the organizations that deploy technology and pair it with expert human recovery capabilities will capture revenue that others leave on the table.
MD Clarity combines an AI-powered revenue optimization platform with human-powered revenue recovery services to help provider organizations reclaim the revenue they have already earned. Its RevFind platform digitizes payer contracts and automatically validates every payment at the line-item level, while its team of payer reimbursement specialists manages the appeals, follow-ups, and escalations required to convert identified underpayments into collected revenue. From multi-state MSOs to independent practices, MD Clarity clients are recovering millions in revenue that legacy systems and manual processes consistently miss.
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.
