Thank you dear subscribers, we are overwhelmed with your response.
Your Turn is a unique section from ThePrint featuring points of view from its subscribers. If you are a subscriber, have a point of view, please send it to us. If not, do subscribe here: https://theprint.in/subscribe/
Recently there was a national emergency in India. No, the country’s sovereignty was not under threat. It was simply that India, despite its achievements in nuclear technology, missile development and space exploration, still struggles to conduct high-stakes examinations without controversy.
The latest crisis involved the 2026 NEET-UG examination. Conducted by the National Testing Agency for more than 2.27 million aspiring medical students, the examination had to be cancelled after investigators found significant overlap between a leaked “guess paper” and the actual question paper. Several arrests followed, including individuals associated with the examination process itself. Such was the seriousness of the situation that the Defence Forces were reportedly called upon to assist in conducting the re-examination.
The incident raises an obvious question. In an era of secure digital payments, satellite control and artificial intelligence, why do examination systems remain vulnerable to paper leaks?
A national emergency of a different kind
Paper leaks have become an unfortunate feature of recruitment examinations, entrance tests and public service selections. Their consequences extend far beyond cancelling a single examination. Millions of students spend months, often years, preparing while families make enormous sacrifices. When a paper is compromised, it is not merely an examination that fails; public confidence in fairness itself suffers.
The traditional examination process creates multiple points of vulnerability. Questions are prepared weeks in advance, printed, transported, stored and distributed to examination centres. Every stage involves people, and every human interface creates an opportunity for compromise. The larger the examination, the greater the risk.
Lessons from an open-book examination
This controversy reminded me of a very different experience from the early 1990s at the School of Artillery.
Computer education had only recently been introduced into the curriculum, and I happened to be responsible for conducting some of those courses. One aspect of our examination system always puzzled me. We seemed to believe that examinations were primarily tests of memory rather than tests of understanding.
As an experiment, I allowed students to carry textbooks into the examination hall. Many were surprised.
Yet it was easy to design questions that could not be answered by copying from a book. Students still needed to understand the subject, identify relevant information, interpret it correctly and apply it to solve practical problems. In many ways, the examination became a better measure of competence.
The experience taught me an enduring lesson. The real challenge in assessment is not preventing access to information. It is designing questions that evaluate understanding, judgement and application.
AI can plug the leaks
Artificial intelligence now offers an opportunity to address another long-standing weakness of examinations.
Instead of preparing question papers weeks in advance, a secure AI-assisted system could generate the final paper only minutes before the examination begins. Drawing from an expert-curated question bank, the system could assemble a paper matching predefined syllabus coverage and difficulty.
The completed paper could then be encrypted and transmitted simultaneously to examination centres through secure digital channels. The decryption key would be released only a few minutes before printing locally at authorised centres.
Such a system would dramatically reduce opportunities for leakage because the lengthy chain of printing, transportation and storage—the weakest links in today’s process—would simply disappear.
The concept can go further.
Adjacent candidates could receive different papers while being assessed against identical standards. Organised paper leaks and mass copying would become significantly more difficult because there would no longer be a single examination paper to steal.
AI can also evaluate question banks, detect duplication and bias, ensure syllabus balance and maintain comparable standards. After the examination, it can analyse response patterns to identify statistical anomalies that may indicate organised malpractice.
Technology is not the constraint
None of this eliminates the need for human oversight. Robust cybersecurity, independent audits, secure digital infrastructure and expert academic supervision would remain indispensable. Question banks would require constant review, and AI systems themselves would need rigorous validation.
No examination system will ever be completely foolproof.
However, reducing the window for malpractice from several weeks to a matter of minutes represents a transformational improvement. It replaces blind trust with verifiable processes and greatly narrows opportunities for organised fraud.
India has repeatedly demonstrated its ability to build digital infrastructure at an extraordinary scale. Whether through digital payments, identity systems or space technology, the country has shown that ambitious technological solutions can succeed when backed by administrative will.
The integrity of examinations deserves similar attention.
Ultimately, the larger lesson extends beyond technology. Examinations should measure understanding rather than memory, and assessment systems should rely more on secure design than on trust alone. AI cannot eliminate dishonesty, but it can make it far more difficult while strengthening fairness and merit.
For a nation where the future of millions of young people depends upon a single examination, that would be a reform well worth pursuing.
(The author is an Indian Army veteran and a contemporary affairs commentator.)
These pieces are being published as they have been received – they have not been edited/fact-checked by ThePrint.