NHA detects fraudulent claims worth ₹582.11 crore in PM-JAY

New Delhi: The National Health Authority’s (NHA) has identified fraudulent claims worth ₹582.11 crore under Ayushman Bharat PM-JAY since the scheme's launch in 2021.

Ayushman Bharat PM-JAY is the world's biggest health assurance scheme for the poor, providing free health cover of up to ₹5 lakh per family per year.

NHA’s National Anti-Fraud Unit (NAFU) has deployed 62 trigger-based techniques such as application of Artificial Intelligence (AI)/Machine Learning (ML), facial comparison NLP (Natural Language Processing), etc to detect potential fraud and safeguard financial transactions from unethical practices.

Also Read | Centre to train joint directors for Ayushman Bharat Digital Mission

“These triggers have enabled NAFU to identify confirmed fraudulent claims worth ₹582.11 crore under Ayushman Bharat PM-JAY. The system demonstrates an average trigger efficacy of approximately 45%, indicating a robust fraud detection mechanism,” the health ministry annual report 2024-25 said.

Using Optical Character Recognition (OCR) and NLP technologies, the system has flagged 28,000 claims as suspicious under Ayushman Bharat PM-JAY. Of these, 14,000 claims have been confirmed as fraudulent, with a total fraud value of ₹31 crore.

Facial comparison technology has identified 2,790 claims as suspicious. Of these, 119 claims amounting to ₹18 lakh have been confirmed as fraudulent.

Also Read | Odisha becomes latest state in the country to implement Ayushman Bharat PM-JAY

Machine learning (ML) and artificial intelligence (AI) techniques have flagged 90,000 claims as suspicious. Among these, 510 claims have been confirmed as fraudulent, amounting to a total value of ₹40 lakh.

The rule-based analysis feature focuses on analyzing structured transactional data to identify fraudulent activities with precision. It identified 410,000 suspicious claims of which 1.49 lakh claims have been confirmed as fraudulent, amounting to ₹232.39 crore.

Also Read | Centre’s telemedicine service integrated with Ayushman Bharat Digital Mission

According to the government, to uphold accountability, severe punitive actions have been enforced, including the de-empanelment of 1,080 hospitals, suspension of 689 hospitals, and filing of 20 FIRs against violators.

Additionally, penalties totaling ₹125.03 crore have been imposed on erring hospitals for various violations.

Database tampering

The ₹562.4 crore worth of false claims hints at deeper issues—possibly misuse of API endpoints or database tampering by empanelled hospitals using tactics like SQL injection or unauthorized system access. Although AI-based fraud detection has been introduced, it's likely being outsmarted by methods like synthetic identity fraud and automated evasion.

The National Anti-Fraud Unit (NAFU) has rolled out 57 advanced tools, including machine learning and fuzzy logic, but their success depends on real-time detection and clean training data. Without strong security layers like end-to-end encryption, mandatory multi-factor authentication, and regular penetration testing, the system remains vulnerable to manipulation and insider threats," said Shashank Shekhar, co-founder, Future Crime Research Foundation (FCRF).