New Delhi : The Ahmedabad Police Cyber Crime Branch on Wednesday arrested four men who allegedly used artificial intelligence, including Google’s Gemini AI tool, to create deepfake videos of a businessman and bypass Aadhaar biometric verification to carry out a fraudulent loan transaction without triggering OTP alerts.
The case came to light after a city-based businessman, who is engaged in import-export work, noticed he had stopped receiving OTPs from his bank for two days and approached the police suspecting something was wrong.
During the investigation, the police found that the mobile number linked to his Aadhaar had been changed without OTP verification and that his biometric data had been tampered with.
The investigators also found that a bank account was also opened in his name and a loan of Rs 25,000 was taken from it. Police also discovered that the gang also accessed businessman’s DigiLocker account in a bid to retrieve stored documents.
According to the officials, the gang used Google’s Gemini AI tool to generate deepfake videos of the victim, which were then used to bypass Aadhaar biometric authentication and change his registered mobile number, redirecting all OTPs to a number controlled by the accused.
Using the businessman’s Aadhaar number and other personal data, the gang attempted to open accounts across three banks through e-KYC. They succeeded with Jio Payments Bank, through which they took a Rs 25,000 loan.
One of the arrested men worked at a Common Service Centre and allegedly misused his access to Aadhaar systems and official kits to execute the mobile number change.
Investigators said they were surprised by the sophistication of the method used in the fraud.
The four arrested have been identified as Kanubhai Parmar, Ashish Vanand, Mohammad Kaif Patel and Deep Gupta.
The incident comes days after Finance Minister Nirmala Sitharaman chaired a high-level meeting with top bank officials to discuss how rapidly advancing artificial intelligence tools could pose a risk to the country’s banking system.
The discussion centred around the potential cybersecurity threats linked to Anthropic’s Claude Mythos model, a powerful AI system that has recently come under scrutiny after reports of unauthorised access.
