Nilekani Panel Explores Last-Minute Question Papers to Prevent Exam Leaks

A government task force headed by Infosys co-founder Nandan Nilekani is exploring a system in which examination question papers are generated at the last minute from a large digital question bank to reduce the risk of paper leaks. The panel is also considering wider computer-based testing (CBT), which would require expanding India’s network of secure examination centres.

The proposals form part of a broader examination-reform exercise launched after the NEET-UG paper leak controversy. Speaking at the Bloomberg New Economy Forum in New Delhi on October 13, 2026, Nilekani argued that preparing a fixed question paper well before an examination creates an incentive for people to obtain and sell it illegally. He suggested that technology, including artificial intelligence (AI), could help generate papers closer to the time of the test. The ideas are under consideration and should not be treated as a confirmed nationwide policy.

Key Takeaways

• The Nilekani-led task force is examining last-minute generation of question papers from a large question bank.
• AI could help create question papers on the day of an examination, according to Nilekani’s remarks.
• The panel is also considering a wider transition to computer-based testing.
• India has around 1.5 lakh locations where examinations can be conducted, but additional capacity and secure infrastructure would be required for a nationwide expansion.
• The examination-reform task force was formed following the NEET-UG paper leak controversy.
• The panel’s recommendations are expected within three months of its formation on July 28, 2026.
• Digital testing could reduce some risks associated with printing and transporting physical papers, but cybersecurity, question quality, accessibility and operational reliability remain important challenges.

What Is the Nilekani Panel Proposing?

The central idea under consideration is to reduce the time between the creation of an examination paper and the actual test. Instead of preparing one fixed paper well in advance and distributing it through a chain of people and institutions, examination authorities could maintain a large, secure bank of questions and generate the final paper shortly before candidates begin the examination.

Nilekani raised the question of why an examination needs to depend on a single, pre-prepared question paper when the paper itself can become an object of illegal trade. His proposed approach is to reduce the opportunity for anyone to obtain the final set of questions ahead of time.

The panel is examining technology as part of the solution, including the possibility of AI-generated question papers. The idea is not simply to use a chatbot to write an examination. A workable system would need a structured question bank, controlled access, clear syllabus coverage, verified answers, difficulty calibration and processes to ensure that the generated paper meets the examination’s requirements.

The task force is still developing its recommendations. No final implementation schedule or nationwide rollout of AI-generated examination papers has been confirmed in the cited reports.

How Last-Minute Question Paper Generation Could Work

A digital question bank would contain a large collection of questions mapped to the relevant syllabus, topics, learning outcomes and difficulty levels. Questions would need to be reviewed, validated and categorised before they could be used in a high-stakes examination.

Shortly before a test, an authorised system could select or generate a paper according to pre-defined rules. Those rules might specify the number of questions from each subject area, the distribution of difficulty levels, the marks assigned to each section and the time required to complete the examination.

AI could potentially assist with generating new questions, adapting existing questions and assembling papers that meet a set of specifications. Human subject experts would still be needed to validate accuracy, remove ambiguity, check for duplicate or overly similar questions and ensure that the paper measures the intended knowledge and skills.

The final paper could then be delivered through a secure digital testing system or prepared for controlled printing, depending on the examination’s format. The key security objective would be to prevent the complete final paper from being available to unauthorised people well before the examination begins.

A possible digital examination workflow

1. Build a question bank: Collect and categorise questions across the syllabus.
2. Review and validate: Have qualified subject experts check correctness, clarity and difficulty.
3. Set examination rules: Define subject coverage, question count, marks and difficulty distribution.
4. Generate the paper close to the test: Use a controlled system to select or create the final questions shortly before the examination.
5. Secure delivery: Restrict access and distribute the paper through protected systems or controlled printing.
6. Audit the process: Maintain logs of access, paper generation, changes and delivery.

This is an illustrative workflow, not a confirmed blueprint released by the task force. The actual design would depend on the recommendations, examination format and technical capacity of the authorities responsible for implementation.

Why Exam Paper Leaks Remain a Serious Problem

Question paper leaks can undermine the fairness of competitive examinations by giving some candidates access to information that others do not have. In high-stakes tests, an advance copy can have substantial value because admission, recruitment and career opportunities may depend on a candidate’s score.

The risk is not confined to the person who originally prepares a paper. A paper may pass through multiple stages involving question setters, reviewers, printers, storage facilities, transport arrangements, examination authorities and other service providers. Each additional point of access can create another potential security weakness.

Once a paper is leaked, authorities may face difficult choices. They may need to investigate the breach, determine how widely the questions circulated, cancel the examination or conduct a re-test. Such decisions can impose financial costs, delay admissions and create uncertainty for students and families.

The NEET-UG controversy in 2026 intensified calls for structural reform. The controversy led to a major disruption in the medical entrance examination process and student protests, after which the government constituted a high-powered task force to examine how public examinations are designed, secured and administered.

The panel’s approach aims to address the incentives and processes that can make leaks possible, rather than relying exclusively on investigations after a breach has occurred.

How AI Could Help Generate Examination Papers

AI could play several roles in an examination system. It may help draft questions from approved content, create variations of existing questions, classify questions by topic, flag possible duplication and support the assembly of papers that satisfy defined rules.

For example, an examination authority could specify that a test must contain a fixed number of questions from each subject area, with a defined balance between easy, moderate and difficult questions. AI-assisted tools could help identify suitable questions from a validated bank or propose new ones for expert review.

However, AI-generated questions are not automatically accurate or fair. A system can produce incorrect answers, ambiguous wording, questions outside the prescribed syllabus or items that are unexpectedly difficult. It may also generate questions that appear different but test substantially the same concept.

These risks are particularly important in examinations such as NEET-UG, where small differences in scores can affect admission outcomes. Question quality, syllabus alignment and consistency across different versions of a paper are therefore essential.

A credible system would need human oversight, independent validation, documented rules and testing before deployment. AI should assist the examination process without becoming an unchecked authority over the content of high-stakes assessments.

Maintaining fairness between different question papers

If candidates receive different papers or question combinations, the examination authority must ensure that the versions are comparable in difficulty and coverage. A paper that happens to contain substantially easier questions could create an unfair advantage.

Authorities could use a large bank of validated questions, statistical analysis and carefully designed paper-setting rules to improve comparability. The precise methods would depend on the examination and the extent to which papers differ between candidates.

For high-stakes examinations, fairness requires more than generating questions quickly. It requires evidence that the questions measure the same abilities and that scores can be interpreted consistently across test versions.

Computer-Based Testing Is Another Major Focus

Alongside last-minute paper generation, Nilekani said the task force is examining the possibility of moving examinations towards computer-based testing.

In CBT, candidates answer questions on computers at designated examination centres. The system can deliver questions digitally, record responses electronically and reduce the need to print, package and transport large volumes of paper.

Digital delivery can also make it easier to administer different versions of an examination, subject to rules that maintain comparable difficulty and content coverage. It can reduce some physical handling risks and provide digital records that may support audits and investigations.

However, CBT is not automatically secure. Examination platforms can face cyberattacks, technical failures, unauthorised access, impersonation attempts and problems involving local networks or devices. Secure software, identity checks, access controls, monitoring and contingency plans would remain essential.

The system also needs to handle the practical needs of millions of candidates. A national transition would require sufficient computers, reliable power, secure networks, trained staff and alternative arrangements for centres affected by technical problems.

India Needs More Examination Centres

Nilekani noted that India has approximately 1.5 lakh locations where tests can be conducted. A substantial expansion of computer-based testing would require strengthening and increasing this network.

A larger centre network could reduce the pressure on existing facilities and make it easier to schedule examinations across locations. It could also reduce the need to concentrate large numbers of candidates at a limited number of sites.

Yet the number of locations alone does not determine capacity. Each centre must have enough functional computers, appropriate seating, reliable electricity, secure connectivity and staff trained to manage the examination. Centres also need to meet accessibility requirements and provide consistent conditions for candidates.

Geographical distribution is another consideration. Students in remote and rural areas should not face disproportionately long travel times or higher costs because a test is available only at a distant computer centre. Expansion plans would need to account for regional access, transport links and local infrastructure.

The government and examination authorities would also need to decide how centres are selected, audited and monitored. Outsourcing parts of the testing infrastructure can expand capacity, but clear accountability and consistent security standards are necessary.

What Happens to Printed Examinations?

The panel’s proposals raise a broader question about whether every high-stakes examination needs to rely on a fixed paper prepared well in advance.

Generating the paper shortly before a test could reduce the period in which the final questions are exposed to potential leaks. If a test is delivered digitally, the system could also avoid some of the physical processes involved in printing and transporting papers.

But not all examinations can immediately move to computer-based formats. Some tests require handwritten responses, diagrams, practical work or other formats that may be difficult to reproduce digitally. Infrastructure constraints may also make a nationwide transition challenging in the short term.

A phased approach could be more practical, with different examinations adopting digital delivery according to their requirements and the readiness of the testing network. Any transition would need to protect accessibility and ensure that candidates are not disadvantaged by unfamiliar technology.

The panel has not announced a final decision to eliminate printed question papers across all examinations. Its current exploration should be understood as part of a wider review of examination security and delivery.

The Task Force’s Mandate and Timeline

The government constituted the high-powered examination-reform task force under Nilekani on July 28, 2026, following the NEET-UG paper leak controversy. It was tasked with examining systemic issues in the conduct, security, design and governance of public examinations.

The panel is expected to submit its recommendations within three months of its formation. The report is therefore a key milestone for understanding which proposals may move beyond discussion into formal recommendations.

Nilekani has said the prime minister is keen to implement the panel’s suggestions. That statement indicates political interest in the reform process, but it does not mean that every proposal has already been approved or that a particular implementation date has been set.

The report is expected to clarify the task force’s broader approach, including the role of technology, examination-centre capacity, question-setting procedures and governance. The authorities will then need to determine how the recommendations can be implemented, funded and monitored.

Challenges That the Reforms Must Address

The proposed model could reduce certain risks, but its effectiveness would depend on how it is designed and managed.

Cybersecurity: A digital question bank can become a valuable target. Authorities would need strong authentication, restricted access, encryption, monitoring and incident-response procedures.

Question quality: AI-generated content must be checked for factual accuracy, syllabus alignment, ambiguity and unintended difficulty differences.

Fairness: Different papers must remain comparable in subject coverage and difficulty. Candidates should not receive materially different assessments simply because a system generates questions at random.

Infrastructure: A wider CBT network requires computers, power, secure software, connectivity, maintenance and trained personnel.

Accessibility: Candidates with disabilities and those with limited familiarity with digital systems may require suitable accommodations and support.

Governance and accountability: Clear responsibility is needed for approving questions, managing the question bank, investigating breaches and deciding what happens if an examination is disrupted.

Contingency planning: Digital examinations need procedures for power failures, network outages, software errors and other interruptions that could affect candidates.

These considerations show why technology should be part of a broader reform programme rather than treated as a standalone solution.

What the Proposal Could Mean for Students

For students, the most important potential benefit is a more secure examination process. If last-minute generation and secure delivery work as intended, it could become harder for individuals to obtain a complete final question paper in advance.

A stronger CBT network could also make examinations more flexible to administer. However, students may need time to become familiar with digital testing interfaces, and authorities would need to provide clear guidance and practice opportunities before high-stakes examinations move to new formats.

The change would also raise questions about how results are calculated and how comparable performance remains across different papers and sessions. Students and teachers would need transparent information about question formats, marking rules, technical arrangements and grievance procedures.

Until the task force releases its recommendations and examination authorities announce specific changes, candidates should not assume that upcoming examinations will immediately adopt AI-generated papers or nationwide CBT.

The Bigger Picture

The Nilekani panel’s proposals reflect a shift in how India could approach examination security: instead of depending primarily on keeping one fixed paper secret for an extended period, authorities could use secure question banks, late-stage paper generation and digital delivery to reduce the opportunity for leaks.

The approach could help close vulnerabilities in the examination process, but its success will depend on the quality of the technology and the institutions operating it. AI must be paired with expert review, comparable question difficulty, robust cybersecurity, reliable centres and clear accountability. A secure examination system must protect both the confidentiality of questions and the fairness of the assessment.

Looking Ahead

The next major development will be the task force’s report, expected within three months of its formation on July 28, 2026. Its recommendations should help clarify whether last-minute question generation, AI-assisted paper setting and wider computer-based testing will be adopted, and which safeguards would be required. Until an official decision is announced, these remain proposals under consideration rather than confirmed changes to the examination system.

If the recommendations are accepted, implementation will require coordination among examination authorities, technology providers, centre operators and subject experts. The government will need to expand infrastructure, establish common security standards, validate digital assessment tools and provide candidates with clear information. The long-term test will be whether the reforms reduce leaks while preserving reliable, accessible and fair examinations for students across India.

Frequently Asked Questions

1. What has Nandan Nilekani proposed to prevent exam paper leaks?

Nilekani has suggested maintaining a large bank of questions and generating the final examination paper at the last minute. The idea is to reduce the opportunity for people to obtain and sell a fixed paper before the test.

2. Will AI generate NEET question papers?

AI-generated question papers are being explored as part of the broader reform discussion. No final nationwide policy confirming that NEET-UG papers will be generated by AI has been announced in the cited reports.

3. What is computer-based testing?

Computer-based testing allows candidates to take examinations on computers at designated centres. It can reduce the need for printed papers and enable digital delivery, but it requires secure software, reliable infrastructure and strong operational safeguards.

4. When will the Nilekani panel submit its report?

The task force was formed on July 28, 2026, and is expected to submit recommendations within three months of its formation. The report will provide more clarity on the proposals and possible next steps.

5. Will printed question papers be discontinued immediately?

No such nationwide decision has been confirmed in the cited reports. The panel is examining alternatives, and any change would depend on the final recommendations, government decisions and the readiness of examination infrastructure.