Exam fraud ring used AI-blended photos to bypass biometric checks in IBPS Clerk test
Police in Lucknow arrested 10 members of an organised fraud ring that used AI-based face-mixing software to blend photos of proxy candidates with genuine IBPS exam applicants, allowing imposters to bypass biometric identity checks. The gang, operating across multiple Indian states, created forged identity documents to support the deception until invigilators detected suspicious behaviour at an exam centre.
What happened
A 10-member organised fraud ring in Lucknow used AI-based apps to blend photographs of real candidates and proxy test-takers until they appeared nearly identical, matching approximately 70% of facial features. The gang created forged Aadhaar and PAN cards to support the fraudulent identities, allowing proxies to enter exam centres posing as genuine IBPS Clerk Examination 2025 candidates.
Police arrested the gang members including alleged mastermind Anand Kumar, 34, an assistant manager at Prathama UP Gramin Bank in Sambhal, from the Bijnor road area in Lucknow on October 7, 2025. Invigilators at a Bijnor exam centre grew suspicious of a candidate's behaviour, leading to the arrest of proxy Abhishek Kumar and the unravelling of the inter-state network.
Reported impact
The fraud ring operated across Bihar, Uttarakhand, and Uttar Pradesh, with two suspects remaining at large at the time of reporting. Police recovered 16 mobile phones, 21 ATM cards, nine Aadhaar cards, seven PAN cards, multiple laptops, and Rs 1.53 lakh in cash from the arrested members.
Index analysis
The gang used AI-based face-mixing software to circumvent biometric and visual identity verification checks at exam centres. The technology allowed proxy candidates to pass screening by achieving high levels of facial feature similarity to the registered candidates they impersonated.
Governance controls
Exam centre invigilators detected suspicious behaviour during the IBPS Clerk Examination, initiating the investigation that exposed the fraud ring within 48 hours. The incident revealed a gap in identity verification procedures that allowed AI-modified images to pass initial screening checks.
Classification
Sources and evidence
Sources not yet published
This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.