How UIDAI is using AI to tackle payment frauds

Source: The post is based on the article “How UIDAI is using AI to tackle payment frauds”  published in Indian Express on 3rd August 2023.

What is the News?

The Unique Identification Authority of India(UIDAI) is using an artificial intelligence-based system called Finger Minutiae Record – Finger Image Record (FMR-FIR) to address fraud in the Aadhaar-enabled Payment System (AePS).

What is Finger Minutiae Record – Finger Image Record (FMR-FIR) Technology?

Finger Minutiae Record – Finger Image Record (FMR-FIR) is an Artificial Intelligence /Machine Learning technology used by UIDAI to prevent fraud in Aadhaar-enabled Payment System (AePS).

This technology is able to check the liveness of a fingerprint to detect the use of a cloned fingerprint during the authentication process.

It uses a combination of both finger minutiae and finger image to verify if a fingerprint is from a real or cloned finger.

What is the need for UIDAI to use this FMR-FIR Technology?

This step was taken due to cases of fraudulent silicone-based fake fingerprints which were used to steal money, especially affecting the rural AePS user base.

According to the Home Ministry, there were 2.62 lakhs financial crimes reported in the fiscal year 2020-21, which includes money laundering, bribery, corruption, and various frauds. This number significantly increased to 6.94 lakhs in 2022.

Can technology alone solve the fraud problem?

While technology can be helpful in combating financial frauds, it has limitations.For example, it has not been able to prevent certain AePS-related frauds where business correspondents (BCs) in the payment supply chain are involved. 

A business correspondent(BC) is an informal bank agent with a biometric Point-of-Sale (PoS) machine similar to a micro ATM. 

To receive money, individuals provide their bank details and Aadhaar-based biometrics to the BC.Unfortunately, some BCs manipulate the transaction amounts by entering higher values in their system.This often goes unnoticed by unsuspecting individuals, particularly in rural areas, who may not ask for receipts that the BC is supposed to generate after each transaction.