The National Testing Agency (NTA) is looking to deploy artificial intelligence (AI) in question paper setting to reduce human involvement and lower the risk of examination-related breaches, Director General Abhishek Singh said on Saturday.
“One usage of AI that we do very often is how do you ensure that you develop a larger bank (of questions) and then reduce the human element in setting question papers,” Singh said, speaking at an event on AI and digital public infrastructure (DPI) organised by the National Council of Applied Economic Research (NCAER).
The NTA’s examination processes have remained under intense scrutiny since the NEET-UG 2024 controversy, when allegations of a question paper leak, irregularities in the award of grace marks, and an unusually high number of top scorers triggered nationwide protests and a Supreme Court-monitored investigation. Fresh concerns resurfaced this year after the NTA cancelled NEET-UG 2026 last month following allegations that the question paper had been leaked, prompting court proceedings, demands for institutional reforms, and renewed debate over the security of high-stakes entrance examinations.
Singh said the recent controversy stemmed largely from excessive reliance on human processes and that greater use of AI and technology could help reduce such vulnerabilities.
“The recent controversy which we are battling with primarily came because there was too much of a human element. When one human has exposure to a larger number of questions which might be asked and if that human goes rogue and is compromised, then all these challenges come in,” said Abhishek Singh.
“So, if we use technology, if we use AI to distribute that and ensure that there is a lot of tech element — and less of human element, which will still be required in some way — then you reduce the probability of anything going wrong,” Singh added.
Singh said the growing use of AI in education could eventually transform the way students are assessed by reducing the emphasis on rote learning and limiting the role of coaching institutions that rely on memorisation-based preparation, while enabling examinations that better evaluate students’ reasoning, critical thinking, problem-solving abilities, and potential to contribute to research and innovation.
“In education, the use of technology and AI will ultimately go to that level wherein we will be eliminating rote learning and eliminating the ability of coaching institutions and these learning schools or whatever, shops or factories wherein they make the children memorise everything and then they go and write it out in three hours without actually being able to test his ability to reason, actual ability to think and the actual ability to contribute to research,” Singh said.
The comments come amid ongoing efforts to strengthen the integrity of entrance examinations. The agency has since initiated a series of reforms, including tighter security protocols and greater use of technology, while policymakers have increasingly advocated the use of AI and digital tools to strengthen exam integrity and shift assessments away from memorisation-based learning towards testing higher-order cognitive skills.
Singh said the NTA is exploring the use of AI for translating question papers across the multiple languages in which its examinations are conducted, while ensuring that the process takes place in secure, air-gapped environments rather than conventional cloud-based systems to prevent sensitive examination content from being exposed.
Additionally, he said AI and data analytics could help the agency make better use of examination telemetry data, such as the time students take to answer individual questions, to more accurately assess question difficulty and candidate ability. Over time, this could enable a shift towards computer-adaptive testing, where candidates receive questions of varying difficulty levels based on their performance rather than all students attempting the same question paper.
Singh also noted that India’s research funding is about 0.6 per cent of GDP, while China’s is at 3 per cent. India needs to invest more in research and innovation as several Indian scholars currently go abroad and develop intellectual property there, he said.
He also highlighted that while the government has several initiatives, such as the Research, Development and Innovation (RDI) Fund, private sector participation remains low.
“Investments need to come from our TCSes, Infosyses, and our Jios who are sitting on a lot of capital,” he added.
