It is 4 am somewhere in rural Bihar. A 17-year-old girl — the first in her family to sit for NEET — has not slept. She has memorised cellular respiration diagrams, practised stoichiometry problems, sacrificed two years of adolescence to a single morning. She believes, with the faith of the truly deserving, that the examination tomorrow will be fair. That belief, in 2024, was broken. The Supreme Court of India recorded it as an undisputed fact: the NEET-UG question paper, sat by over 23.3 lakh candidates across 4,750 centres in 571 cities, was leaked. The girl’s two years were weighed against a criminal network’s single night.
This is not a story about one leak. It is a story about a structural flaw so fundamental that the State itself admitted it had no clean solution. Cancel the exam? Twenty-three lakh innocent students punished. Let it stand? Merit corrupted. The Supreme Court, in an act of judicial anguish, chose the lesser injustice. No democracy should ever be forced to make that choice.
The Architecture of Betrayal
Every leaked examination follows the same anatomy. A question paper is set months in advance, printed, packaged, transported, stored, and distributed. Each step — printing press, transport vehicle, district warehouse, examination centre strong-room — is a separate point of catastrophic failure. Security cameras, sealed envelopes, biometric checks: all defend individual links in a chain whose weakest link is sufficient to compromise every candidate in the country, regardless of measures everywhere else. This is a textbook weakest-link vulnerability — a concept well-understood in cybersecurity, ignored in examination design.
Parliament responded with the Public Examinations (Prevention of Unfair Means) Act, 2024 — assented on February 12, 2024, brought into force on 21 June 2024 — imposing imprisonment of up to ten years and fines of ₹1 crore for organised cheating syndicates. A necessary measure. But punishment is retrospective. It operates after the harm is done, after twenty-three lakh futures have been gambled with. Can we not adopt a system meant for preventive intervention? An architecture in which there is, quite simply, no paper left to leak?
The Solution: Engineering Truth into the System
I would propose a National Dynamic Examination Question Bank (NDEQB): a centralised, cloud-vaulted repository of over 100,000 verified, expert-created questions. Instead of a single paper compiled months in advance, an AI-driven assembly engine generates a unique paper for every candidate, moments before the examination begins.
‘Psychometrically balanced’ is the engineering constraint, not a marketing phrase: each paper differs in its specific questions but is statistically equivalent in difficulty level, discrimination index, and overall fairness — techniques already standard in international standardised testing. No candidate is advantaged or disadvantaged by which subset they receive.
Proposed End-to-End Zero-Trust System Architecture for India’s Public Examinations
Five Walls That Cannot Be Breached Together
The mechanism can be built in such a manner that it does not rely on a single wall. It can have five independent, overlapping secu
rity layers — a defence-in-depth model used in nuclear command systems and critical national infrastructure worldwide.
>> Layer 1 requires simultaneous authorisation from the Ministry of Education, a designated cyber security agency, and an independent auditor to access the question vault — neutralising the insider-collusion model behind virtually every major leak investigated to date.
>> Layer 2 transmits questions via post-quantum encrypted micro-tunnels: no PDFs, no printouts, no static file that can be photographed or smuggled, because no such file is ever created.
>> Layer 3 decrypts questions only inside a hardware-isolated processor cache at the terminal, with local storage and OS access blocked entirely.
>> Layer 4 renders each paper for one-time viewing and permanently purges it from memory upon submission.
>> Layer 5 embeds an invisible pixel-level watermark encoding candidate identity, seat location, and timestamp in every rendered screen: if any image escapes the first four layers, it can be traced to the precise source — candidate, seat, and second — within minutes.
A Social Justice Instrument Dressed as a Security Protocol
There is a second injustice running parallel to paper leaks, quieter and more systemic: the coaching industry.
Today, an entire commercial ecosystem exists to predict “important questions,” sell pattern-analysis packages, and offer privileged access to test-preparation resources that cost more than many Indian families earn annually. A student in a metropolitan coaching hub purchases a measurable advantage over an equally capable student in a government school in Chhattisgarh. That advantage has nothing to do with subject mastery. It is a purchasable artifact of a static, predictable examination structure.
When the question pool exceeds 100,000 items and every paper is uniquely assembled, the return on memorising predicted questions collapses to near zero. What becomes valuable is the one thing that should always have been the currency of merit: actually knowing the subject. The system I am proposing does not merely secure examinations. It restructures the incentive architecture of Indian education — aligning it, for the first time, with what the National Education Policy 2020 explicitly demands: conceptual understanding over rote memorisation. A security architecture becomes, simultaneously, a social justice instrument, levelling the field between the child of a daily-wage labourer and the child of privilege on the morning that matters most.
The Constitutional Imperative
My proposal asks the state for nothing new. It asks the state to give technological effect to commitments the Constitution has already made. Article 14 guarantees equality before law — a system where access to a leaked paper confers advantage on a subset of candidates is, in substance, a violation. Article 16 guarantees equality of opportunity in public employment — recruitment examinations compromised by leaks directly undermine this guarantee for every honest candidate. Articles 21 and 21A, as interpreted by Indian courts, recognise that a fair assessment process is integral to the right to a dignified livelihood, particularly for candidates without access to private coaching infrastructure. The proposed architecture does not merely comply with these guarantees. It operationalises them.
The choice before India is no longer between perfect security and imperfect security. It is between continuing to patch a structurally leakable model, year after year — or building, once, an architecture in which there is nothing left to leak.
For NEET, JEE, UPSC, SSC, Railways, Banking, and every state-level recruitment examination, the proposed model offers a credible, scientifically grounded, legally anchored path to an examination ecosystem that is secure, equitable, and genuinely meritocratic. The 17-year-old in rural Bihar deserves no less. Neither does the Constitution that promised her equality.
(Views expressed by the columnist are personal and do not necessarily reflect the opinion or policy of the news portal)