CMS Deputy Administrator and Chief Operating Officer Kim Brandt speaks during the Feb. 18 Fed Tech Priorities Summit. Zaid Hamid/Nextgov/FCW

The Centers for Medicare and Medicaid Services is eyeing a combination of artificial intelligence, medical provider education and new laws to reduce improper payments and fraud.

One of the biggest drivers of improper payments — which totaled billions of dollars across CMS programs in fiscal year 2025 — is incorrect documentation or a lack of documentation, according to CMS Deputy Administrator Kimberly Brandt.

“Our goal is to really start using machine learning tools and some of those technologies to really help start working with the provider community,” Brandt said at the Health Datapalooza event on Friday.

Brandt’s office did not clarify how the agency plans to use AI to support providers by the time of publication. But she stressed that “if we can reduce that documentation improper payment, that would drop our entire improper payment rate down by, at least in Medicare, as much as two percentage points, which is huge.” 

CMS is also working with contractors to replace legacy Medicare claims systems with a modern one as part of its fraud crackdown. The new claims system would issue “real-time payments” to providers and include “real-time fraud review tools,” Brandt said.

CMS selected two companies to compete for a potential seven-year contract for the modern Medicare claims system, Washington Technology previously reported. HealthEdge’s contract had an initial $2.5 million obligation and a $1.1 billion ceiling, while Peraton’s contract had an initial $9.1 million obligation and an $825 million ceiling, the outlet reported.

Brandt’s office did not respond to a request for comment on the claims systems overhaul competition by the time of publication.

CMS has selected Tegria, a health care consulting and technology services company, to provide management and technical implementation support for the claims systems overhaul.

The agency already uses many technologies to curtail fraud, waste and abuse in government programs. 

In March 2025, CMS launched a Fraud Defense Operations Center to proactively detect fraud. The center uses AI to sift through thousands of Medicare claims and create a heat map of frequently billed items and services.

“Skin [substitutes] were a big thing last year. We’re still seeing the overflow of catheters [and] genetic testing. But the fact is that we decided those are the areas that we know that we can hit hard at,” Brandt said.

Within those identified items and services, CMS uses machine learning to find documentation patterns that would benefit from human scrutiny, Brandt explained.

The center has stopped nearly $2.5 billion in potentially fraudulent Medicare payments and taken action on just under 800 providers, Brandt said Friday.

She previously told Nextgov/FCW that CMS would benefit from new dashboards and better analytics to detect fraud in state-administered Medicaid programs.

Besides technology, Brandt sees a role for Congress to play in helping CMS crack down on fraud. 

She said she has urged congressional offices to give CMS the power to sometimes stop providers and payers from enrolling in federal programs.

“If we can keep them out, it’ll be better for all of us,” Brandt said.