Google Pay Deploys AI-Driven Teams to Combat Fraudulent Transactions

279
11 Jul 2024
5 min read

News Synopsis

Google Pay is leveraging advanced artificial intelligence (AI) and machine learning (ML) technologies to enhance its ability to detect suspicious digital transactions, addressing the alarming rise in payments fraud in India.

Surge in Digital Payments Fraud

Digital payments fraud in India surged over fivefold to a record ₹1,457 crore in FY24 from the previous year, according to data from the Reserve Bank of India (RBI).

Google's Commitment to Security

"Keeping users safe online is critical for us. We have a dedicated team working on building fraud detection technology at Google Pay and invest heavily in network and software security research to ensure our users can trust the app," said Sharath Bulusu, Director of Product Management at Google Pay.

Market Share and Transaction Volume

Google Pay, which holds the second-largest market share of 35 percent in the Unified Payments Interface (UPI) ecosystem (based on transaction value) after PhonePe, processed over 5 billion transactions in April.

Successful Fraud Prevention

In 2023, Google Pay reportedly thwarted approximately ₹12,000 crore worth of suspicious transactions, including 1 lakh suspicious transactions declined every day. The app has developed an in-house early detection system, DigiKavach, which analyzes transaction patterns in real-time.

Early Detection System: DigiKavach

Launched in October last year, DigiKavach analyzes various signals such as transaction time, amount, spending patterns, and Know Your Customer (KYC) details to detect anomalies while ensuring user privacy in compliance with the Digital Personal Data Protection Act (DPDP).

User Alerts and Notifications

When a potential fraudulent transaction is detected, users receive alerts in multiple languages, often with bilingual notifications to ensure clarity. For example, if a user suddenly spends a large amount at an unfamiliar merchant, the system flags it and alerts the user to verify the transaction.

Future Initiatives and Collaboration

Bulusu highlighted the potential benefits of upcoming initiatives like the RBI's proposed Payment Intelligence Platform, which aims to enhance fraud detection through standardized data sharing.

The interoperability and open standards that underpin the UPI system have been crucial to its success. Creating a shared platform for fraud detection is expected to benefit all users by providing a safer ecosystem.

Central Bank's Dedicated Platform

On June 27, the central bank announced a dedicated platform to address digital payment fraud in the country. The proposed platform would gather intelligence across the digital payments network and enable real-time data sharing with stakeholders like payment apps.

A committee under AP Hota, former managing director and CEO of the National Payments Corporation of India (NPCI), which operates UPI, is expected to give its recommendations soon.

Collaboration with Key Stakeholders

"Collaboration is critical. We work with DFS, RBI, NPCI, and MeitY to both contribute what we learned because as the operators of a product, there are certain types of insights that we gain firsthand," Bulusu said.

Recent Trends and Innovations

On June 12, representatives from Google Pay, along with private sector banks and NPCI, participated in a meeting organized by the Department of Financial Services (DFS) to discuss recent fraud trends on the app and innovations in response.

Rise in UPI Transactions and Banking Fraud

This dramatic rise in fraud aligns with India's rapid transition to a digital payments hub, driven by the widespread adoption of UPI since 2016. UPI transactions have skyrocketed 137 percent over the past two years to an impressive ₹2 lakh crore, according to RBI data.

The largest number of banking fraud incidents in FY24 were related to the card/internet category, with 29,082 incidents, a 334 percent increase from the previous year. The 29,082 frauds in FY24 accounted for ₹1,457 crore worth of card/internet fraud, the highest in the past six years and a fivefold jump from FY23.

Podcast

TWN Special