In a first-of-its-kind initiative, the Union Public Service Commission (UPSC) used Artificial Intelligence (AI) to identify and reject ineligible applications for the Civil Services (Preliminary) Examination, 2026, at the application stage itself. Through this exercise, UPSC rejected 569 applications, including those submitted by candidates who had exhausted the maximum number of attempts allowed and those involving multiple applications by the same individual.
The move comes two years after the dismissal of former IAS trainee Puja Khedkar from service in 2024. Khedkar was found to have appeared for the Civil Services Examination (CSE) 2022 despite exhausting the maximum number of attempts by allegedly changing her name and her parents’ names. Her candidature was subsequently cancelled by UPSC.
Until last year, such verification checks were conducted only at the interview stage, after candidates had cleared both the preliminary and main examinations. As per UPSC rules, candidates from the general category are allowed six attempts up to the age of 32 years, OBC candidates are allowed nine attempts up to the age of 35 years, while SC and ST candidates can appear an unlimited number of times up to the age of 37 years.
For the Civil Services Examination 2026, a total of 8.18 lakh candidates registered for the examination held on May 24, while 5.49 lakh candidates appeared. The number of applicants was lower than the 9.5 lakh registrations recorded in 2025.
According to UPSC sources, one possible reason for the decline in applications was the introduction of several new measures this year, including Aadhaar-based authentication and AI-driven de-duplication of applications.
Last year, UPSC launched a new application portal that provided candidates the option of Aadhaar authentication for the 2026 examinations. Officials stated that approximately 94 percent of applicants opted for Aadhaar verification, ensuring that they were unique applicants with no possibility of duplication.
For the remaining approximately 49,000 applicants who did not opt for Aadhaar authentication, UPSC used AI to identify possible duplicate applications by analysing details such as names, parents’ names, dates of birth, and photographs.
Following this process, UPSC used AI to examine whether applicants had already exhausted the number of attempts permitted for their category or had crossed the prescribed age limit. The verification was carried out by comparing applications with the Commission’s database covering the last 15 years.
As a result, UPSC identified 569 candidates in the Civil Services Examination who were ineligible due to exceeding the attempt limit or age criteria, leading to the rejection of their applications. Officials also revealed that the same AI-based tool detected 69 ineligible candidates in the Indian Forest Service examination held on May 24.
Commenting on the initiative, UPSC Chairman Ajay Kumar said that technology must serve the cause of fairness. He stated that UPSC undertook the de-duplication exercise at the Preliminary Examination application stage to ensure that every genuine candidate is correctly identified while fraudulent or multiple applications are detected and removed. He added that the exercise relied on secure comparison with the Commission’s existing database while ensuring complete protection of candidates’ information. According to him, UPSC aims to create a level playing field for all candidates, with technology serving as an important enabler in achieving that objective.
Officials also said that UPSC conducted category verification at the Preliminary stage for the first time. Earlier, checks regarding changes in candidates’ social categories from previous attempts were carried out only at the interview stage. Using AI, UPSC examined whether applicants had changed the category selected in earlier attempts, such as moving from General to EWS or from SC to OBC.
Through this exercise, UPSC identified 43,497 candidates whose category details differed from previous applications. The Commission subsequently sent emails to all such candidates to verify the changes. Sources indicated that many candidates had previously applied under the General category because they were unable to obtain the required certificates in time.
The verification process ultimately resulted in the cancellation of 133 applications, which formed part of the total 569 rejected applications, after it was found that the candidates had exceeded the number of attempts permitted under their respective categories.
The AI-driven screening exercise marks a significant procedural change for UPSC, enabling the Commission to identify duplicate and ineligible applications before the examination process progresses further.
Also Read: ₹21,000 Crore Mundra Drug Bust: ED Raids Delhi Businessman, Tracks Terror Funding Links https://www.vibesofindia.com/ed-raids-delhi-businessman-in-rs-21000-crore-mundra-port-heroin-seizure-case/
