AI Helps CAG Identify Fake Beneficiaries Of Government Schemes, Fake Schools Claiming Scholarships

The Comptroller and Auditor General of India (CAG) is making significant use of artificial intelligence (AI) and machine learning (ML) in various areas.

They have been able to identify non-existent schools that falsely claim scholarship benefits and detect instances where multiple beneficiaries of government schemes use similar images.

At the SAI20 Engagement Group Summit, held under India's G20 Presidency in Panaji, the CAG presented several case studies demonstrating the use of AI and ML. These case studies were compiled in a 'Compendium on Responsible Artificial Intelligence'.

One of the showcased case studies focuses on the implementation of AI in detecting duplicate, fake, and ineligible beneficiaries of the government's digital literacy programme, Digital Saksharta Abhiyan (DISHA).

The programme required beneficiaries to upload their photographs, and the CAG used an intelligent model developed with the help of the Centre for Data Management and Analytics (CDMA) to automatically analyse these images. This automated analysis was crucial due to the large volume of images involved.

By utilising an open-source platform called Python, the CAG's CDMA successfully developed a model that could determine whether the same images, different images of the same beneficiaries, or non-human images were used for claiming training costs.

This model enabled the automatic analysis of a significant number of images, which would have been impractical to do manually.

The detection of these cases has been instrumental in identifying risky transactions, duplicate beneficiaries, and fake or ineligible beneficiaries.

In addition, the CAG utilised an AI and ML model to detect non-existent schools that were fraudulently claiming scholarship benefits.

The model, developed in Python, used pre-defined risk parameters from 2019-20 data to identify suspected fake schools claiming scholarships in 2017-18. The model achieved an accuracy rate of over 92 per cent, which helped in identifying samples for further field-level verification.

AI was also utilised by the CAG to uncover ineligible beneficiaries who were fraudulently claiming benefits from the scholarship scheme designed for marginalised communities.

The CAG is also employing AI and ML in the performance audit process of programmes such as PM-Kisan and Pradhan Mantri Awas Yojana (Gramin).

This utilisation of advanced technologies aims to enhance the efficiency and effectiveness of the audit process, ensuring a comprehensive evaluation of these schemes.