New Delhi: The government’s new AI-based fraud-detection system, designed to counter digital payment scams, has prevented potential losses of nearly ₹400 crore within just five months of its launch, officials confirmed.
Developed by the Department of Telecommunications (DoT), the Fraud Risk Indicator (FRI) has so far flagged or blocked more than 6.5 million suspicious transactions, either by generating real-time alerts or declining them automatically.
Launched in May 2025, the FRI uses advanced data analytics to identify risky transactions linked to phishing, impersonation, fake loan schemes, digital arrest rackets, and investment frauds. It categorises mobile numbers into medium, high, and very high risk, identifying nearly 5,000 fraudulent numbers every day.
The system’s rollout comes amid a surge in cybercrime, as India’s digital ecosystem continues to expand rapidly — with over 86% of households now connected to the internet. Reported cybersecurity incidents more than doubled from 1.029 million in 2022 to 2.268 million in 2024, prompting stronger government intervention.
In the 2025–26 Union Budget, ₹782 crore was allocated for cybersecurity measures. Since then, authorities have blocked over 9.42 lakh SIM cards and 2.63 lakh IMEIs linked to cybercriminal networks.
Between May and mid-September, the FRI prevented 4.8 million fraudulent transactions, saving about ₹140 crore. By early October, the figure rose to ₹220 crore, with 5.5 million transactions flagged or stopped. Just a month later, that figure has nearly doubled to ₹400 crore saved.
Authorities have also frozen nearly 3 million bank accounts suspected of being involved in digital scams. The milestone follows Google Pay’s integration with the FRI system in October, which helped prevent ₹90 lakh worth of fraud within just two weeks.
A senior DoT official said the system’s reach is growing rapidly. “While we can’t guarantee complete prevention in every digital arrest case, the FRI’s effectiveness is increasing daily. Paytm alone now generates nearly 30,000 fraud alerts per day, up from 10,000 earlier,” the official added.
