NEET exam modernisation: CBT, AI, and question vaults explained

Following repeated NEET-UG controversies, India's High-Powered Task Force—led by former UIDAI chairman Nandan Nilekani—is studying three major changes to how the exam is conducted: shifting from pen-and-paper to computer-based testing (CBT), creating a secure digital vault of pre-vetted questions, and potentially using artificial intelligence to assemble question papers at the last minute. While these measures aim to prevent paper leaks and improve exam security, they also introduce new risks—from digital infrastructure failures to fairness concerns for rural and less digitally-experienced candidates. Understanding what these changes mean is crucial for NEET aspirants planning their preparation and exam strategy.

Why NEET is considering a complete overhaul

The push for modernisation stems from two major NEET crises. In 2024, widespread paper leaks and irregularities forced a national re-examination. Then in 2026, NEET-UG was cancelled entirely and re-conducted across 5,440 centres on June 21, 2026, affecting over 22.79 lakh students. These back-to-back failures eroded public confidence and prompted the government to form a High-Powered Task Force to fundamentally rethink how the exam is designed and delivered.

The Dr. Radhakrishnan Committee had already recommended 101 reforms after the 2024 fiasco, but many were implemented without pilot studies or evidence-based validation. The new Task Force is now examining whether technology—specifically CBT, secure question management, and AI—can address the root causes: paper leaks, insider involvement, and logistical vulnerabilities that plague a pen-and-paper exam involving millions of candidates across thousands of centres.

What is computer-based testing (CBT) and how would it change NEET?

Computer-based testing means NEET would be conducted entirely on computers at examination centres, rather than on printed paper with optical mark recognition (OMR) sheets. The NTA already uses CBT for many other exams, and the shift is intended to eliminate risks tied to printing, physical transportation, and OMR tampering—all vulnerabilities that have been exploited in past NEET leaks.

However, CBT is not a silver bullet. It removes physical paper risks but introduces digital ones: server failures, cybersecurity breaches, software vulnerabilities, and examination-centre technical glitches. More importantly, CBT does not address the core institutional problem—insider involvement and compromised networks within the NTA itself, which have been alleged in past controversies. A digital exam is only as secure as the people and systems managing it.

The equity and access challenge: who gets left behind?

One of the most serious concerns with CBT is the digital divide. Rural candidates, students from schools with limited computer access, and those less familiar with typing or navigating digital interfaces may face a disadvantage compared to urban, privileged peers. A candidate who is slower at using a computer or unfamiliar with digital navigation could lose valuable time during the exam—not because they lack medical knowledge, but because of a skill gap unrelated to the subject.

Technical failures also pose a fairness risk. System crashes, internet outages, and server problems have been reported even at examination centres in major cities. If your exam is disrupted by a technical glitch, you may be disadvantaged through no fault of your own. The Task Force has not yet clarified how such incidents would be handled or whether affected candidates would receive compensation or a re-attempt.

The question vault: a secure pool of pre-vetted questions

Instead of preparing a single question paper months in advance (which gives time for leaks), the proposed system would maintain a large, continuously updated digital vault of vetted questions. For each NEET session, questions would be randomly selected from this vault and assembled into a paper just before the exam, then securely transmitted to centres. The idea is to minimise the window during which a final paper could be leaked, copied, or sold.

To work fairly, this vault would need to contain thousands of questions covering all syllabus units, question types, and difficulty levels. Every question would require expert review, independent solving, answer-key verification, and translation into multiple languages. Papers across different sessions would need to be comparable in difficulty—a major challenge when you're running 16–20 separate NEET sessions to accommodate 24 lakh candidates.

The vault itself is not leak-proof. A large repository of sensitive questions is a bigger target for insider theft. Past NEET controversies have shown that compromised insiders and networks can access confidential material regardless of where it is stored. The security of a question vault depends entirely on robust access controls, clear separation of responsibilities, and independent oversight—none of which can be guaranteed by technology alone.

Artificial intelligence in question paper assembly: promise and peril

The Task Force is exploring whether AI could help assemble question papers from the vault or even generate new questions. In theory, AI could speed up the process and reduce human involvement—and thus reduce opportunities for insider manipulation. However, AI-based systems introduce their own risks: prompts and outputs could be intercepted, data could be exposed, and AI-generated questions might have accuracy or translation errors that a human expert would catch.

Importantly, the Task Force has not yet formalised AI's exact role. The focus is on last-minute assembly from a pre-vetted question pool using conventional software—a simpler, lower-risk approach than AI generation. Any AI component would be studied separately and only adopted after rigorous testing. The key message: AI is being considered as a tool, not a solution, and its use depends on evidence that it improves security without compromising fairness or accuracy.

The normalisation problem: are scores from different sessions truly comparable?

If NEET runs in multiple sessions with different question papers, a critical problem emerges: how do you compare scores fairly? A candidate who gets 600 marks in Session 1 may have faced a harder paper than a candidate who scored 580 in Session 3. To address this, the NTA would likely use statistical normalisation—converting raw marks into percentile scores so that performance is comparable across sessions.

However, normalisation is not perfect. It relies on the assumption that question papers are truly comparable in difficulty, syllabus coverage, and discriminatory power—an assumption that is hard to guarantee, especially if AI is involved. Normalisation also cannot eliminate all uncertainty. In NEET, where even a small change in rank can determine whether you get into a premier medical college or not, even modest adjustments to normalised scores can affect thousands of candidates near the admission cutoff.

Before nationwide adoption, the NTA must demonstrate through rigorous pilots and independent validation that papers are comparable and that the normalisation method is transparent, defensible, and does not disadvantage candidates based on which session they appear in. This is not a technical detail—it is a fairness issue that directly affects your admission prospects.

What do global exams like SAT and GRE do differently?

International standardised tests like the SAT and GRE use CBT and have sophisticated question-development processes: expert review, pretesting, statistical analysis to assess difficulty and bias, and rigorous validation. However, these exams serve a different purpose than NEET. The SAT and GRE assess broader academic readiness and skills; NEET is a highly competitive selection exam where scores directly determine medical admission. With millions competing for a limited number of seats, the stakes are far higher in India.

China's Gaokao, by contrast, remains largely paper-based despite being a similarly high-stakes competitive exam. The point is not that one format is universally better, but that any system adopted for NEET must be tailored to Indian conditions: the enormous candidate pool, language diversity, intense competition, and the critical importance of question accuracy and fairness. Global practices offer lessons, but they cannot be simply copied.

What happens next: pilots, validation, and a phased rollout

The Task Force is not proposing an immediate switch to CBT or AI-based systems. Instead, the plan is to conduct controlled pilots, undergo independent audits, and gather evidence before any nationwide implementation. This is the right approach: technology should be tested under realistic Indian conditions before it affects the admission prospects of millions of students.

Before you sit for NEET, the NTA must demonstrate that CBT reduces security risks without introducing new ones, that question vaults and AI systems are reliable and fair, and that normalisation across sessions is transparent and defensible. Robust safeguards, clear accountability, and contingency plans must be in place. The test of modernisation is not technological sophistication, but demonstrable gains in security, reliability, and fairness.

What this means for your NEET preparation right now

If you are preparing for NEET in 2026 or 2027, the exam format you face depends on when the Task Force completes its pilots and validation. You may still encounter pen-and-paper NEET, or you may be among the first cohorts to take a computer-based version. Either way, your preparation strategy should remain the same: master the syllabus, practise with high-quality mock tests, and build speed and accuracy.

If CBT is introduced, familiarise yourself with typing and navigating digital interfaces well before the exam. Practise mock tests on computers, not just on paper. Understand that technical glitches can happen and that the NTA will need to have clear protocols for handling them. Most importantly, stay informed: follow official NTA announcements and MedAdmit updates as the Task Force releases its findings and recommendations.

Frequently asked questions

The modernisation of NEET through CBT, question vaults, and AI is still in the planning and pilot phase. Your immediate focus should remain on mastering the syllabus and building exam skills, regardless of format. Stay updated with official NTA announcements and MedAdmit's coverage as the Task Force releases its findings—the exam you take may look very different from today's NEET.