Blab AI
Indiana Launches 90-Day AI Pilot to Spot Medicaid Fraud
Indiana’s Medicaid system is putting artificial intelligence to the test in a 90‑day experiment aimed at catching fraud before it reaches the state’s wallets. The Family and Social Services Administration (FSSA), in partnership with the U.S. Centers for Medicare & Medicaid Services (CMS) and Texas‑based Oracle, will use the state’s own Medicaid data to spot suspicious billing patterns, halt improper payments, and hand investigators ready‑made case packages.
Oracle’s platform is engineered to flag claims in real time, so regulators can intervene before a payment is made. The system is not a replacement for FSSA staff and will not launch enforcement actions on its own.
"Our goal is to make sure every dollar is spent correctly and to keep Medicaid programs sustainable," said FSSA Deputy Chief of Staff Marcus Barlow. He added that the AI tools will scrutinize both providers and the more than one in four Indiana residents who rely on Medicaid.
To protect privacy, the data‑use agreement signed by Oracle requires the state to strip all personal and protected health information before it is shared. Oracle must destroy any restricted data it receives and is barred from sharing FSSA data with external parties.
Oracle began its analysis on Wednesday morning, Barlow said. The state’s Medicaid spending has surged—from roughly $2.1 billion in 2017 to nearly $5 billion for fiscal year 2027, a 134 % rise over ten years.
Since Governor Mike Braun took office, Indiana has tightened the Healthy Indiana Plan with new work requirements, capped the autism therapy known as Applied Behavior Analysis, and ordered some providers to stop advertising Medicaid programs. The FSSA has also launched an "expanded oversight" period, adding 400 new employees to manage increased eligibility checks for the Healthy Indiana Plan.
The pilot is intended to evaluate the challenges of deploying AI in this context and to determine whether a similar model could be rolled out to other states. After the 90‑day period, CMS will review a proof‑of‑concept report to guide future strategies and investments.
Barlow expressed enthusiasm about the results. "We’re excited to see how this technology can improve program integrity and reduce fraud, waste, and abuse," he said.
The initiative follows a national trend of using AI to analyze health‑care claims. CMS has been encouraging states to adopt advanced analytics for fraud detection, and other states have begun similar pilots.
If the system proves effective, it could shape future federal guidance on AI use in Medicaid and become a model for other states seeking stronger oversight while safeguarding patient privacy.
At this time, no specific performance metrics or cost‑savings estimates have been released. CMS will report on the pilot’s outcomes after the 90‑day period.
In short, Indiana is testing an AI system that could flag Medicaid fraud in real time, potentially cutting improper payments and enhancing program sustainability. The program is overseen by FSSA, funded by CMS, and implemented by Oracle, with strict data‑privacy safeguards in place. The results will determine whether similar AI tools can be scaled across the country.
Oracle’s platform is engineered to flag claims in real time, so regulators can intervene before a payment is made. The system is not a replacement for FSSA staff and will not launch enforcement actions on its own.
"Our goal is to make sure every dollar is spent correctly and to keep Medicaid programs sustainable," said FSSA Deputy Chief of Staff Marcus Barlow. He added that the AI tools will scrutinize both providers and the more than one in four Indiana residents who rely on Medicaid.
To protect privacy, the data‑use agreement signed by Oracle requires the state to strip all personal and protected health information before it is shared. Oracle must destroy any restricted data it receives and is barred from sharing FSSA data with external parties.
Oracle began its analysis on Wednesday morning, Barlow said. The state’s Medicaid spending has surged—from roughly $2.1 billion in 2017 to nearly $5 billion for fiscal year 2027, a 134 % rise over ten years.
Since Governor Mike Braun took office, Indiana has tightened the Healthy Indiana Plan with new work requirements, capped the autism therapy known as Applied Behavior Analysis, and ordered some providers to stop advertising Medicaid programs. The FSSA has also launched an "expanded oversight" period, adding 400 new employees to manage increased eligibility checks for the Healthy Indiana Plan.
The pilot is intended to evaluate the challenges of deploying AI in this context and to determine whether a similar model could be rolled out to other states. After the 90‑day period, CMS will review a proof‑of‑concept report to guide future strategies and investments.
Barlow expressed enthusiasm about the results. "We’re excited to see how this technology can improve program integrity and reduce fraud, waste, and abuse," he said.
The initiative follows a national trend of using AI to analyze health‑care claims. CMS has been encouraging states to adopt advanced analytics for fraud detection, and other states have begun similar pilots.
If the system proves effective, it could shape future federal guidance on AI use in Medicaid and become a model for other states seeking stronger oversight while safeguarding patient privacy.
At this time, no specific performance metrics or cost‑savings estimates have been released. CMS will report on the pilot’s outcomes after the 90‑day period.
In short, Indiana is testing an AI system that could flag Medicaid fraud in real time, potentially cutting improper payments and enhancing program sustainability. The program is overseen by FSSA, funded by CMS, and implemented by Oracle, with strict data‑privacy safeguards in place. The results will determine whether similar AI tools can be scaled across the country.
