HHS Deploys AI to Analyze State Health Audits for Fraud
The Department of Health and Human Services announced Thursday it will use ChatGPT and other AI tools to systematically review audit reports from all 50 states. The initiative aims to identify fraud patterns and reduce improper spending across federally funded health programs.
Gustav Chiarello, HHS assistant secretary for financial resources, said the department receives thousands of audit reports annually that previously went largely unexamined. "It's classic big government: Everyone files an audit and it lands with a thud and no one does anything about it," Chiarello said. "Here, with AI, we're able to dig into it."
The program will analyze audits from state Medicaid programs, federal research grants, addiction services, and other HHS-funded initiatives. Any organization receiving at least $1 million annually in federal health dollars must submit these reports.
How the System Works
Generative AI tools can identify patterns across large document sets that human reviewers might miss. HHS said it is analyzing existing public audit reports rather than generating new investigative data.
States and grantees that fail to file required audits or ignore identified problems face potential loss of federal funding. HHS sent formal notices to all 50 states stating it will no longer accept indefinite informal resolution of audit noncompliance or material weaknesses.
Concerns About Accuracy and Bias
Critics raised questions about the reliability of AI-driven enforcement. Chiarello acknowledged the department is evaluating reports rather than uncovering original evidence, but did not detail specific safeguards against AI errors.
Rob Weissman, co-president of Public Citizen, questioned the administration's intentions. "The AI is kind of beside the point when you assess what their actual objectives are, rather than what they pretend they are," he said.
The administration has previously acknowledged major mistakes in data used to justify fraud investigations. At least one New York Medicaid probe relied on flawed data, according to documents reviewed by the Associated Press.
Broader Anti-Fraud Push
This initiative builds on existing HHS efforts to automate fraud detection across Medicaid and Medicare. Federal Trade Commission Chairman Andrew Ferguson recently cited AI tools flagging suspected fraud in student loan programs and other areas.
Chiarello said he is encouraging other federal departments to adopt similar AI-based audit analysis systems. "It would be fairly easy for the other agencies to use our technology and jump on it," he said.
Your membership also unlocks:
