Artificial intelligence (AI) can support the question-making process for national entrance exams, but it cannot independently replace human expertise, IIT Madras director Veezhinathan Kamakoti has said, stressing that peer review remains indispensable at all stages of exam preparation.Talking about the evolving role of AI in exam systems, Kamakoti said AI could generate potential questions, but those questions would still require rigorous scrutiny by subject matter experts before they could be used in an actual exam.“I don’t think AI can independently set a question paper. AI may be able to generate questions, but someone still needs to review them,” he said.His comments come at a time when educational institutions and examination bodies across the country are increasingly exploring AI-powered tools for academic and administrative functions. However, Kamakoti warned that relying solely on AI for a national entrance exam would not be appropriate.
Human supervision essential in multilingual exams
To highlight how difficult it was for the National Eligibility and Entrance Test (NEET), Kamakoti pointed out that the test is conducted in 12 different languages. He mentioned that quality control was necessary in such situations because translating the questions from one language to another should not change their meaning.“NEET is very inclusive in that sense. It is conducted in 12 languages, so someone has to verify the translations as well. AI can certainly help with translation, but someone has to scrutinize the translations,” he said.The comments highlight one of the biggest challenges in India’s exam ecosystem, maintaining consistency and accuracy across multiple languages while ensuring fairness for millions of candidates.
AI can complement, not replace, exam experts
According to Kamakoti’s observation, AI should not be considered a replacement for the academic experience but instead should be considered an enabling technology. Although AI systems can be used to facilitate some activities such as question generation, translation and content processing, decision-making on issues such as question quality, difficulty level, factual and linguistic accuracy should remain in the hands of academic and specialist experts.Furthermore, his observation is consistent with the general philosophy of examination reform, where the use of technology aims to improve efficiency and transparency while ensuring academic standards and the integrity of national entrance examinations.Given the growing use of AI in education, Kamakoti’s comments are pertinent and relevant, especially for assessments such as those that determine admission to professional programs and impact the careers of millions of students annually.