The Comptroller and Auditor General of India (CAG) is embracing the power of Artificial Intelligence (AI) and Machine Learning (ML) in its audit practices, with a focus on identifying fraud and ensuring accountability. During the SAI20 Engagement Group Summit, held under India’s G20 Presidency, the CAG unveiled a “Compendium on Responsible Artificial Intelligence,” which featured several case studies showcasing the application of these technologies.
One notable case study highlighted the use of AI in detecting duplicate, fake, and ineligible beneficiaries of the Digital Saksharta Abhiyan (DISHA), a government program focused on digital literacy. By leveraging an intelligent model developed using Python, the CAG’s data management and analytics center automatically analyzed a large volume of beneficiary photographs, identifying instances where the same images were used multiple times or non-human images were submitted. These findings helped uncover risky transactions, duplicate beneficiaries, and ineligible claims.
Another case study demonstrated the use of AI and ML to identify non-existent schools falsely claiming scholarship benefits. By developing a machine learning model in Python, the CAG was able to detect suspected fake schools based on pre-defined risk parameters. The model achieved an accuracy rate of over 92%, aiding in the identification of potentially fraudulent cases for further field-level verification.
The CAG also utilized AI to detect ineligible beneficiaries attempting to claim benefits under welfare schemes aimed at marginalized communities. By leveraging AI algorithms, the CAG successfully identified circular trading transactions in taxation, where fake invoices are used to claim input tax credit. These applications of AI and ML have enhanced the efficiency and accuracy of auditing processes, enabling the CAG to uncover irregularities and fraudulent activities more effectively.
The implementation of AI and ML by the CAG extends beyond sampling and field validation. These technologies have also been utilized in performance audits of pre-matric and post-matric schemes, utilizing data available on the National Scholarship portal. The preliminary analysis has been shared with the government, and the final report is expected to be released soon.
The adoption of AI and ML by the CAG signifies a significant step toward leveraging advanced technologies to enhance audit practices, promote transparency, and safeguard public funds. By harnessing the capabilities of AI and ML, the CAG aims to strengthen its ability to detect fraud, identify anomalies, and ensure the efficient implementation of government programs.
