CMS Launches AI Challenge to Combat Medicare Fraud

What's Happening?

The Centers for Medicare & Medicaid Services (CMS) has initiated the Crushing Fraud Chili Cook-Off Competition, a market-based research challenge aimed at combating Medicare fraud using explainable artificial intelligence (AI) and machine learning (ML) models. The competition seeks to identify anomalies in Medicare Fee-for-Service (FFS) claims data that can be translated into indicators of fraud. CMS is focusing on scalable technologies that reduce labor-intensive processes while ensuring human oversight for effective interpretability. The challenge is designed to enhance the detection of fraudulent activities such as false billing and upcoding, which divert resources away from patient care and undermine public trust.

Why It's Important?

Medicare serves as a crucial healthcare lifeline for millions of Americans, but its complexity makes it susceptible to fraud. By employing innovative, data-driven approaches like explainable AI/ML, CMS aims to proactively identify suspicious behavior and protect public resources. Explainable AI/ML techniques are critical as they provide transparency and accessibility to program integrity teams, regulators, and policymakers, fostering trust and enabling fair enforcement actions. The competition's focus on scalable solutions aims to uncover broader patterns and systemic vulnerabilities, allowing for efficient allocation of enforcement resources and maximizing oversight efforts.

What's Next?

The competition will unfold in two phases. Phase 1 involves inviting research proposals tailored to Medicare FFS claims data, with ten teams selected as finalists. The submission period for Phase 1 is from August 19 to September 19, 2025. In Phase 2, finalists will access claims data for a random 5% sample of Medicare beneficiaries and apply their proposed AI/ML techniques. Submissions for Phase 2 are due by December 1, 2025. The winning team will be announced by CMS via social media the week of December 15, 2025.