CMS Uses Explainable AI to Strengthen Medicare Fraud Detection
CMS will use Milliman’s explainable AI model to help investigators focus on high-risk Medicare claims and reduce false-positive fraud alerts.
A new AI-driven fraud detection tool could help the Centers for Medicare and Medicaid Services move faster and smarter.
CMS crowned Milliman as the winner of its Crushing Fraud Chili Cook-Off Competition earlier this month for its “glass-box” AI tool that proactively detects fraud, waste and abuse in Medicare claims, promising new levels of transparency and efficiency for program integrity teams.
Milliman Data Scientist Adam Hearn told GovCIO Media & Research in an interview that traditional fraud detection algorithms often rely on deep learning or neural networks that cannot explain their reasoning and may lack the clinical context needed to assess the severity of a patient’s illness, which can result in false positives for fraudulent claims.
“A provider may appear to have high costs or anomalous behavior, and that could be due to them just managing a sicker population,” said Hearn. “If an AI doesn’t understand that this provider treats sicker, more costly patients, then it could potentially generate false positives that end up wasting investigative resources.”
The glass-box method, unlike traditional AI models, uses a deterministic algorithm rooted in actuarial science and statistics that can be deconstructed to show exactly why a provider was flagged, reducing the time it takes to identify fraudulent Medicare claims while cutting false positives that drain investigative resources.
The tool combines behavioral, network and financial anomalies for all providers in CMS’ limited data sets to generate a single risk score, which investigators can then use to prioritize providers with both high billing costs and statistically abnormal patterns.
“This allows CMS to focus on the providers that are billing thousands or millions of dollars, so when CMS deploys their investigative resources, they’re on the providers that have the biggest financial threat,” said Hearn.
The competition aligns with CMS’ broader push to deploy innovative technologies to combat fraud, waste and abuse. CMS CIO Patrick Newbold said at GovCIO Media & Research’s 2025 Health IT Summit in September 2025 that industry partnerships are essential to improving government efficiency and transparency.
“We have to provide better service for the public and we want industry to actually bring solutions that solve the problems,” Newbold said.
This is a carousel with manually rotating slides. Use Next and Previous buttons to navigate or jump to a slide with the slide dots
-
OPM Extends Tech Force Deadline as Program Tests New Federal Hiring Model
High demand prompted OPM to extend Tech Force applications to Feb. 2 as the agency pilots centralized, skills-based hiring.
5m read -
FDA Outlines AI Principles for Drug Development
New FDA guidance outlines 10 principles for using AI in drug research, development and manufacturing, developed with European regulators.
3m read -
Federal AI Enters a ‘Storming Phase’ Under the Genesis Mission
National labs are working to reduce complexity, connect compute and deploy assured autonomy to speed AI research and deployment.
3m read -
Agencies Turn to Industry to Scale AI, Develop Federal Workforce
Federal IT leaders say industry partnerships are critical to overcoming workforce and technical challenges in AI adoption.
3m read