Daijiworld Media Network
Bengaluru, March 23: In a groundbreaking move to curb exam fraud, the Karnataka Examinations Authority (KEA) has successfully rolled out its AI-powered mobile-based candidate verification system during a high-stakes recruitment exam.
For the first time ever, the system verified candidates' identities using facial recognition technology, ensuring only legitimate test-takers entered the examination hall.
With this pioneering initiative, the KEA has taken a bold step towards eliminating fake candidates, making the recruitment process more secure and transparent.

Representational image
The technology, developed in-house by KEA’s engineering team, was used during a four-day exam for vacant posts, including computer operators and other positions, in the Legislative Council. As candidates entered the exam hall, their faces were scanned using a mobile app, which instantly connected to KEA’s server.
The system then compared the newly captured image with the one submitted by candidates during the application process, providing a real-time verification of their identity.
"This system has successfully prevented fake candidates from entering the exam hall, marking a major breakthrough in ensuring fairness," said H. Prasanna, executive director of KEA. He further explained that the new technology significantly strengthens the examination process, making it foolproof and transparent.
The first day of this technology trial saw 74 candidates take the exam in the morning and 267 in the afternoon, all of whom were verified through the new AI system. Following the completion of the exam, candidates’ OMR sheets were uploaded to the KEA website immediately, further boosting transparency.
"This system opens up the process for anyone to check the OMR sheets, ensuring there are no opportunities for malpractice or doubts," Prasanna added.
With this successful pilot, KEA plans to expand the use of AI-based candidate verification to other upcoming exams, marking a major leap in modernising and securing the recruitment process in the state.