Milliman Wins CMS Fraud Detection Competition, Showcasing Explainable AI
Event summary
- Milliman won the CMS “Crushing Fraud Chili Cook-Off” competition, announced December 23, 2025.
- The winning solution utilizes explainable AI (XAI) to identify fraud, waste, and abuse (FWA) in Medicare fee-for-service claims data.
- Milliman’s tool generates a composite risk score for providers, enabling prioritization of investigations and identification of coordinated fraud networks.
- The competition involved ten finalists who received access to CMS data to apply their AI techniques.
The big picture
The CMS competition highlights a growing trend toward leveraging AI and actuarial science to combat fraud and improve efficiency in public sector programs. This initiative signals a shift from reactive fraud detection to proactive prevention, potentially reducing substantial losses within Medicare and Medicaid. Milliman’s win underscores the value of explainable AI, which is increasingly critical for ensuring accountability and transparency in AI-driven decision-making within government.
What we're watching
- Deployment Scale
- The speed at which CMS integrates Milliman’s solution across its broader programs will determine the tangible impact on fraud reduction and program integrity.
- Competitive Response
- Other actuarial and AI firms will likely accelerate their development of XAI solutions for public sector applications, intensifying competition for government contracts.
- Regulatory Scrutiny
- Increased reliance on AI in government programs will likely draw greater regulatory scrutiny regarding algorithmic bias and transparency, potentially impacting Milliman’s future contracts.
