
HHS Using AI Initiative to Uncover Health Care Fraud, Waste
Key Takeaways
- HHS will apply AI to identify longitudinal, high-dimensional fraud signatures across claims, fill histories, prescriber behavior, and patient populations, rather than relying on isolated anomalous transactions.
- AERO will review at least five years of audits for high-spend grantees, escalating consequences from corrective action demands to withheld payments or grant termination for unresolved findings.
The AI-led initiative was designed to conduct audits and ensure appropriate distribution of health care funds for federal programs.
The US Department of Health and Human Services (HHS) recently announced an initiative backed by artificial intelligence (AI) to review audits and improve oversight in the organization’s move to decrease fraud and waste among federal health programs. Known as the Audit Enforcement and Risk Oversight (AERO) program, AI technology will now be used to detect fraud through records across all 50 US states, according to US News & World Report.1
“Fraud in health care can be defined by its patterns. AI can enable a dynamic, high-dimensional approach to detecting such patterns of potential fraud,” Ravi Patel, PharmD, former innovation lead at the Pittsburgh School of Pharmacy and practicing pharmacist at Blueberry Pharmacy, told Drug Topics®. “A single claim is unlikely to reveal potential fraud on its own. Rather, a longitudinal, aggregate, or composite pattern can expose fraudulent activity through data such as fill histories, prescriber profiles, and patient populations spanning hundreds, thousands, or millions of records.”
The AERO initiative aims to review at least 5 years of audit records for HHS-funded programs, focusing on non-federal entities such as local governments and nonprofits that spend at least $1 million annually in federal funds. This aggressive oversight comes as HHS reveals that hundreds of grantees are years behind on required audits.1
READ MORE:
For pharmacists, the implications are significant, as pharmacies are frequently listed among the provider types convicted in health care fraud schemes, alongside physicians and diagnostic labs. Although the administration intends to work collaboratively with states, it has warned that recipients failing to resolve audit issues could face withheld payments or total grant termination.1,2
The Overall Prominence of Health Care Fraud
The scale of the problem is immense, with some studies estimating that total annual waste in the US health care system ranges from $760 billion to $935 billion, or roughly 25% of total health care spending. Within this figure, fraud and abuse alone account for an estimated $58.5 billion to $83.9 billion annually, according to JAMA.3
Data from the US Sentencing Commission underscores the severity of these crimes. In the 2025 fiscal year, health care fraud cases saw a median loss of over $1.2 million, with nearly 70% of those convicted receiving prison sentences. Notably, 33% of these sentences were increased because the individual abused a position of public trust or used a special skill—a category that often includes licensed health care professionals like pharmacists.4
Pharmacists must be particularly vigilant regarding emerging trends in pharmacy-related fraud, such as those involving opioid prescriptions and pharmacogenetic testing. Drug diversion, which results in the loss of millions of medication doses annually, is often prosecuted under the False Claims Act, as stated in the Journal of the American Pharmacists Association.5
Additionally, the rise of unapproved pharmacogenetic testing kits has created new opportunities for fraudsters to target older adults for their Medicare identification numbers, leading to identity theft and improper billing. Managed care organizations and pharmacists are increasingly using states’ prescription drug monitoring and lock-in programs to flag pharmacy and doctor shopping patterns to curb such abuse.5,6
The Government’s Approach to Reversing Fraud
“The Centers for Medicare & Medicaid Services has repeatedly characterized its evolving approach as a shift away from retrospective investigation toward earlier detection and more stringent enrollment screening,” continued Patel. “While the precise role of AI within either the retrospective or prospective process may be ambiguous, what can be inferred is that the benefits and challenges of AI can apply to fraud in equal measure.”
The implementation of AI through AERO aligns with a broader federal trend of using advanced technology to enhance mission effectiveness. For example, the Department of State has begun leveraging tools like StateChat and Northstar to analyze millions of data points and anticipate global trends.7
However, experts caution that technology must be applied accurately; for instance, the Payment Error Rate Measurement program is often mistaken for a fraud tracker, but it actually measures improper payments—many of which are due to simple documentation errors rather than intentional criminal activity.2
To safeguard their practices, pharmacists are encouraged to conduct periodic self-audits and maintain meticulous records to ensure compliance with the False Claims Act, the Anti-Kickback Statute, and the Stark Law.5 The Academy of Managed Care Pharmacy also advocates for the adoption of electronic prescribing systems to reduce fraud at the pharmacy’s point-of-sale.6
By understanding the patterns that AI is now designed to uncover, ranging from “phantom claims” to illegal kickback loops, pharmacists can better protect their patients and the integrity of the federal health programs they serve.2,5,6
“AI that identifies fraudulent patterns may result in new patterns of fraud. As a result, pharmacy may see a trend toward greater transparency and standardization in documentation, helping practices anticipate the continual retraining and application of AI models,” concluded Patel in his exclusive interview with Drug Topics®. “An ideal future includes thoughtful data governance rules and procedures that enable fraud detection without burdening clinical or operational workflows.”
READ MORE:



















































