On November 11, India observes National Education Day in memory of Maulana Abul Kalam Azad, the country’s first Education Minister. From the 2026–27 academic session, CBSE has introduced computational thinking and artificial intelligence for Classes III to VIII, while children outside school already use AI to write essays and summarise history in seconds. The larger question is what happens to schooling when answers become effortless.
The National Education Policy 2020 called for a shift away from rote learning, and CBSE has since added more competency-based questions to its board examinations. Much of school education, however, still revolves around completing the syllabus and reproducing textbook content in examinations. That imparts essential knowledge but leaves too little room to question, weigh competing claims, or defend a conclusion with evidence. When a machine can summarise a textbook fluently and still be confidently wrong, those limits become harder to ignore.
Judgement is now the scarce resource
AI systems can present contested interpretations as settled facts, omit important perspectives, and invent references that appear genuine. Manipulated videos and digital-arrest scams show how costly misplaced trust in information has become. Finding an answer is easy. Judging whether it deserves belief is harder, and education must therefore teach children to distinguish evidence from assertion, recognise uncertainty, and reach conclusions they can defend.
An older idea of learning
Memory and disciplined learning mattered in Indian intellectual traditions. Nyaya logic, Buddhist and Jain philosophical schools, and debates known as shastrartha also valued reflection, argument, and the examination of opposing positions. One method, purvapaksha, required a scholar to state the opposing view fully and fairly before answering it, and only then to proceed towards a considered conclusion, or siddhanta.
Access to such learning was never universal, but the principle endures. A conclusion earns credibility when it has survived serious scrutiny. In too many classrooms, the balance has tilted towards recall, and AI offers an opportunity to restore it.
The 'Purvapaksha' classroom
Consider a geography lesson on water. Instead of asking students to define rainwater harvesting and list its advantages, a teacher could ask how a fast-growing coastal town should secure its water supply for the next twenty years, choosing among desalination, wastewater recycling, rainwater harvesting, and stricter management of demand. Students could use AI to build the strongest case for each option, weighing cost, energy use, environmental impact, and how the burden is shared, then verify the claims against credible sources and debate the alternatives.
The exercise would end with each student explaining a recommendation, acknowledging its limitations, and identifying what new information might change it. The teacher would guide the discussion, correct errors, and ensure that quieter students were heard. The machine would supply the competing arguments, and the judgement would remain with the students. The method travels across the curriculum. A history class could compare accounts of 1857 and the evidence behind them, a civics class could weigh arguments about a proposed law, and a science class could trace how new evidence displaced an established explanation.
Who decides what children learn?
Curricula need stability, yet they are not immune to politics. Textbooks must be updated as knowledge advances, but India’s successive revisions under governments of different political persuasions show how readily such decisions become contested. Not every revision is politically motivated. The risk lies in selective omission, disproportionate emphasis, and contested interpretations presented as unquestionable fact.
Students should therefore learn to distinguish established facts from interpretations and to ask why credible accounts may differ. A student who can examine an AI-generated answer critically should be able to examine a textbook passage in the same spirit, seeking the evidence and reasoning behind it before accepting or rejecting it. The syllabus should be the foundation of learning and never the boundary of thought. In a democracy, education should prepare citizens to assess competing claims, including claims made by those in authority.
Right answers, best answers, and memory
Not every question has several equally valid answers. The boiling point of pure water at a pressure of one atmosphere and the date of India’s Independence have single correct answers. How a town should manage its water or how a historical figure should be assessed involves evidence, context, values, and trade-offs. Telling the two apart protects students from false certainty where reasonable disagreement exists, and from false doubt where the evidence is conclusive.
Memorisation is not redundant. A student without basic arithmetic, grammar, or a sound knowledge of Indian history cannot reliably evaluate an AI-generated explanation of these subjects. Memorisation should be where inquiry begins rather than where it ends.
What it will take
The change will demand more of teachers, who need subject expertise to lead structured debate and spot fabricated claims. Teacher training must cover the limitations of AI as thoroughly as its operation. Assessment must change too, since classrooms will keep prioritising recall as long as examinations reward it. Questions that require students to weigh evidence and justify a conclusion, backed by short oral assessments, would build on competency-based questions, with careful piloting to ensure fairness.
The opportunity must reach every child. AI tools that work in Indian languages and on limited connectivity will matter most in schools with fewer resources, and the Bhashini platform offers a foundation. Children’s safety must be designed in from the start. The Digital Personal Data Protection Act, 2023, whose provisions on children’s data take effect in May 2027, requires verifiable parental consent and bars the tracking and behavioural monitoring of children, though the Rules give educational institutions a limited exemption. Classroom AI tools should still collect only the data they need and never profile young learners.
Three practical steps could begin the shift. NCERT could develop subject-specific exercises built on the purvapaksha method. CBSE and state boards could pilot assessment formats that reward reasoning alongside foundational knowledge. Teacher-training institutions could make AI literacy, source verification, and the evaluation of competing arguments integral to teacher preparation.
India’s classrooms should keep the syllabus, the textbook, and the discipline of memory, and then go beyond them. The measure of an educated child in the AI age will be how carefully an answer is examined and how soundly a conclusion is defended. Speed has already passed to the machine.
(The author is a former National Cyber Security Coordinator and former Signal Officer-in-Chief of the Indian Army.)
(The opinions expressed in this article are those of the author and do not purport to reflect the opinions or views of THE WEEK.)
