HHS Issues RFI for AI Tools to Stop Healthcare Fraud Before Payments Go Out

HHS Seeks AI Tools to Stop Healthcare Fraud Before Payment

The U.S. Department of Health and Human Services (HHS) and the Centers for Medicare & Medicaid Services (CMS) are moving to use artificial intelligence to flag suspicious claims before money goes out the door. Estimates put Medicare fraud near $60 billion a year, with the HHS Office of Inspector General (HHS-OIG) citing more than $100 billion in improper payments across Medicare and Medicaid in 2023. Many peg total fraud losses at 3%-10% of healthcare spend.

In a February 25, 2026 announcement, Vice President J.D. Vance, HHS Secretary Robert F. Kennedy, Jr., and CMS Administrator Dr. Mehmet Oz outlined steps to move beyond "pay and chase." "For decades, Medicare fraud has drained billions from American taxpayers-that ends now," said Secretary Kennedy. "We are replacing the old 'pay and chase' model with a real-time 'detect and deploy' strategy, using advanced AI tools to identify fraud instantly and stop improper payments before they go out the door."

What HHS/CMS announced

• Deferral of $259.5 million in quarterly federal Medicaid funding to Minnesota while investigations proceed into potentially fraudulent or unsupported claims.
• A nationwide moratorium on Medicare enrollment for select DMEPOS suppliers, a category long linked to high fraud risk.
• A call to action for public input to expand and strengthen fraud prevention, shifting focus to pre-payment detection. "CMS is done trying to catch fraudsters with their hands in the cookie jar-instead, we're padlocking the jar and letting them starve," said Administrator Oz.

Inside the RFI: What CMS wants from stakeholders

HHS and CMS issued a Request for Information seeking ideas to prevent, detect, and respond to fraud, waste, and abuse across Medicare, Medicaid, CHIP, and the Health Insurance Marketplace. Feedback will inform future rulemaking, including a potential "Comprehensive Regulations to Uncover Suspicious Healthcare (CRUSH)" proposed rule.

• Analytics and methodologies that pinpoint risk signals before payment, including AI and data-driven scoring approaches.
• How AI can improve Medicare Advantage coding oversight and hospital billing accuracy at scale.
• Effective AI solution types (including off-the-shelf) for assisting human coders on large record volumes.
• Key features and learning capabilities that raise accuracy, reduce false positives, and prevent errors.
• Lessons learned from real implementations: data needs, model drift, workflow fit, and change management.
• Ways AI can address overpayments and underpayments, and support compliance oversight and audits.

What this means for providers, plans, and revenue cycle leaders

Pre-payment edits, AI-assisted review, and tighter enrollment controls will increase scrutiny on documentation, coding precision, and billing integrity. Teams that prepare now will avoid denials, delays, and enforcement headaches later.

• Map your highest-risk claim types and codes (e.g., DMEPOS, telehealth, high-cost drugs) and validate documentation standards against payer policies.
• Run internal pre-submission checks with rule-based and AI-assisted tools; track false positives and tune workflows for speed and accuracy.
• Evaluate AI vendors for explainability, audit trails, PHI protections, and integration with your EHR/RCM system.
• Tighten MA risk-adjustment controls: verify diagnosis capture, encounter integrity, and supporting notes.
• Stand up an appeals playbook for pre-pay and post-pay reviews with templates, evidence packs, and turnaround SLAs.
• Governance: assign model owners, monitor performance and bias, document training data sources, and log overrides/edits.

If you're modernizing coding and claims review, this resource can help your team upskill quickly: AI Learning Path for Medical Billers.

AI focus areas CMS is probing

• Claims-level anomaly detection that blends code patterns, provider behavior, patient context, and device/supply utilization.
• Document intelligence to cross-check clinical notes, orders, and attachments against codes and medical necessity rules.
• Provider network risk scoring using peer benchmarking, historical actions, billing velocity, and specialty-specific norms.
• Real-time edits to stop suspect claims, paired with clear explanations and pathways for rapid correction.

Guardrails: privacy and appropriate care come first

HHS is clear: fraud prevention cannot compromise beneficiary privacy or block legitimate care. Any AI deployment must protect PHI, minimize access to sensitive data, and maintain medical necessity pathways.

• Privacy and security: minimum-necessary data use, encryption in transit/at rest, role-based access, and detailed audit logs.
• Fairness and accuracy: back-testing for bias, regular calibration, human-in-the-loop for adverse decisions, and quick appeals.
• Transparency: clear reason codes for denials/holds and documented criteria for elevated review.

What to watch next

• The content and timing of a potential CRUSH proposed rule, and how it reshapes pre-payment review and enrollment screening.
• Expansion of moratoria or targeted edits beyond DMEPOS into other high-risk categories.
• State Medicaid program integrity actions mirroring federal moves.
• New HHS-OIG priorities and enforcement trends in the OIG Work Plan.

Bottom line

CMS is shifting fraud control to the front of the payment funnel, with AI as the engine. Healthcare leaders who tighten documentation, deploy explainable AI, and build fast correction paths will protect revenue, reduce risk, and keep patients' care on track.