TRUMP GOVERNMENT DEPLOYS CHATGPT TO READ FIVE YEARS OF MEDICAID AUDITS FROM ALL 50 STATES — HUNTING $200 BILLION IN FRAUD The Department of Health and Human Services announced on May 21, 2026 that it is deploying ChatGPT and other large language models to ingest and analyze audit reports from all 50 states on a rolling basis. The program, called AERO, for Audit Enforcement and Risk Oversight, targets any entity spending more than one million dollars per year in federal funds. The AI tools will scan at least five years of audit history looking for chronic noncompliance, repeat deficiencies, material weaknesses, and delinquent audit obligations. The scale of what is being claimed is remarkable. HHS estimates the federal government loses between 100 billion and 200 billion dollars annually to waste, fraud, and abuse in programs like Medicaid. That range is wide enough to cover the entire annual budget of some countries. The administration is betting that AI can find patterns in audit data that human reviewers have been missing for decades. The enforcement consequences are serious. States and grantees that the AI flags could face temporary payment holds, cost disallowance, suspension or termination of awards, and in the most serious cases, permanent exclusion from federal programs. That last option means losing access to Medicaid funding entirely, which for many state health systems would be catastrophic. HHS Assistant Secretary Gustav Chiarello announced the initiative publicly. It sits alongside Dr. Oz’s broader 50-state Medicaid revalidation effort and the CRUSH program targeting suspicious healthcare billing patterns. The precedent here is significant. This is the first time the federal government has systematically deployed consumer AI tools to audit its own financial relationships with every state in the country simultaneously. Keywords: HHS ChatGPT Medicaid audit, AI government fraud detection, AERO program Medicaid, Trump AI healthcare