AI-Made Question Papers: Nilekani’s Proposal
Why in News?
- On 13 October 2026, Infosys co-founder Nandan Nilekani proposed at the Bloomberg New Economy Forum that question papers be generated at the last minute using AI from a large question bank.
- He chairs the High-Powered Task Force set up after the NEET-UG exam was cancelled in May 2026 following reports of a paper leak.
- He said: "The moment you have a question paper, it becomes an economic incentive for it to be sold."
- He also proposed conducting all examinations through computer-based testing, and said India has 1,50,000 testing locations and needs to expand the network.
- He was asked to submit the Task Force report on 28 July and was given three months to do so.
Key Terminologies
- Question bank
- A large collection of questions prepared in advance. Under the proposal, an AI system would pick or generate a fresh paper from it on the day of the exam.
- Computer-based testing
- An exam taken on a computer at a test centre. Nilekani said all examinations should move to this format.
- High-Powered Task Force
- The body chaired by Nilekani to address exam integrity. An RTI-obtained notification says it was needed because evolving cybersecurity threats were compromising exam integrity.
- Public Examinations (Prevention of Unfair Means) Act, 2024
- The central law notified in June 2024 against paper leaks. It prescribes jail terms of 3 to 5 years and fines for individuals, and larger fines for service providers.
Key Issues
- Leak Incentive: Nilekani argued that a printed question paper has a market value that encourages leaks. His answer is to remove the single paper that can be stolen.
- AI Generation: Papers would be generated on the day from a large bank, so no fixed paper exists before the exam. He said yes when asked if AI could generate a paper on the day of the test.
- Testing Network: He said India has 1,50,000 testing locations and the network must expand. Computer-based testing needs reliable centres, connectivity and invigilation.
- Cyber Threats: The Task Force notification obtained under the RTI Act says rapidly evolving cybersecurity threats were compromising exam integrity. An AI system would itself be a target.
- Earlier Reforms: The 2024 High-Level Committee of Experts, led by former ISRO chairman K Radhakrishnan, made more than 101 recommendations on exam reforms.
Key Implications
Positive/Pros/Merits
- Removes the Prize: If no paper exists before the exam, there is nothing to leak. This addresses the economic incentive Nilekani named.
- Comparable Papers: A large bank can give each candidate a paper of equal difficulty, which is an inference about how such a system would be designed.
- Digital Capacity: A network of 1,50,000 testing locations already exists, which gives a base for computer-based exams.
Negative/Cons/Demerits
- Fairness and Quality: AI-generated papers must be checked for accuracy, difficulty and bias. The proposal does not say how, and this is a gap.
- Infrastructure Gaps: Candidates in remote areas may lack nearby testing locations. The network must expand, as Nilekani himself said.
- Security Risk: A central question bank and generation system becomes a high-value target for hackers, which is an inference.
- Legal Fit: The 2024 Act sets penalties for paper leaks, but the reports reviewed do not say whether it needs amendment for AI-generated papers.
Key Initiatives
- Public Examinations (Prevention of Unfair Means) Act, 2024: Notified in June 2024, it sets 3 to 5 years in jail and a fine up to Rs 10 lakh for using unfair means, and up to Rs 1 crore for service providers.
- High-Level Committee of Experts, 2024: Led by K Radhakrishnan, it made more than 101 recommendations on exam reforms.
- High-Powered Task Force: Chaired by Nandan Nilekani, it was set up after the May 2026 NEET-UG cancellation, and invited suggestions from students and parents in August.
- President Droupadi Murmu, 13 October 2026: She said national security needs close cooperation between scientific institutions and the armed forces, and warned that decisions cannot be handed entirely to AI.
The High-Powered Task Force chaired by Nandan Nilekani was asked to report within three months of 28 July. Nilekani said the single question paper should go and that AI can generate a paper on the day of the test. The reports reviewed do not give the Task Force's final recommendations or any government decision on the proposal.
Director's Perspective
Way Forward
- Pilot AI-generated papers in a low-stakes exam first, with human review of every question set before use.
- Expand computer-based test centres to districts that lack them, before making the format compulsory.
- Audit the question bank and generation system for security by an independent cyber agency.
- Publish the Task Force's recommendations with reasons, so the public can test the proposal.
The idea is bold and aimed at the right target: the leak happens because a fixed paper has a price. AI generation from a large bank could remove that price, and India already has a base of 1,50,000 testing locations. But it shifts risk to the question bank and the software, and the Task Force has not yet reported. In a Mains answer, credit the shift from guarding papers to removing them, then conclude that pilots, audits and wider test centres are needed.
GS Relevance
Frequently Asked Questions
What did Nandan Nilekani propose on 13 October 2026 to stop exam leaks?
Nandan Nilekani proposed generating question papers at the last minute with AI from a large question bank, and conducting all exams by computer-based testing. He said a printed paper creates an economic incentive for leaks, and India has 1,50,000 testing locations that must expand.
What is the High-Powered Task Force on exams?
The High-Powered Task Force, chaired by Nandan Nilekani, was set up after the NEET-UG exam was cancelled in May 2026 following reports of a leak. An RTI-obtained notification says it was needed because evolving cybersecurity threats were compromising exam integrity.
What does the Public Examinations (Prevention of Unfair Means) Act, 2024 provide?
The Act, notified in June 2024, provides 3 to 5 years in jail and a fine up to Rs 10 lakh for using unfair means in public exams. Service providers involved can be fined up to Rs 1 crore, and organised crime attracts a fine of at least Rs 1 crore.
PYQ Practice — Statement Analysis
1 The Public Examinations (Prevention of Unfair Means) Act, 2024 was notified in June 2024.
It prescribes jail terms of 3 to 5 years and fines for using unfair means.
2 Nandan Nilekani proposed that every exam should keep a single fixed question paper printed well in advance.
He proposed generating papers at the last minute from a large question bank, using AI.
3 Computer-based testing was proposed for all examinations by Nilekani on 13 October 2026.
He said India has 1,50,000 testing locations and needs more.