AI in Indian Classrooms: The Revolution Nobody Agreed To
Walk into any Kota coaching centre today and you will find something that would have seemed impossible five years ago: AI tutoring systems generating personalised practice problems at the pace a student can handle, flagging weak areas before a human teacher would notice, and providing doubt-resolution at 2am when the human faculty is asleep. Walk into an urban private school and you will find students using ChatGPT, Gemini, and purpose-built educational AI to draft essays, solve problems, and generate notes from textbooks. Walk into a government school in rural Bihar or Jharkhand and you will find none of this - a technology gap that is being created even as the tools arrive. This is the state of AI in Indian education in 2026: a revolution that nobody explicitly agreed to, affecting students in wildly different ways, with consequences that educators, parents, and policymakers are only beginning to grapple with.
What Is Actually Happening in Indian Schools and Coaching Centres
The penetration of AI tools in Indian education is far deeper than official policy acknowledges. A 2025 survey by the Central Square Foundation found that over 60% of urban high school students regularly use generative AI for homework and assignment preparation. Among JEE and NEET aspirants in Kota, Hyderabad, and Delhi coaching hubs, AI-assisted doubt resolution has become the norm: platforms like Physics Wallah and Unacademy have integrated AI tutoring features that can solve standard problems and explain concepts on demand. Several prominent coaching institutes have deployed AI that analyses a student's test performance to generate targeted practice drills - personalisation at a scale no human faculty could deliver.
The most sophisticated deployments are not in schools at all but in edtech platforms that operate outside the formal education system. These platforms have effectively created a parallel personalised curriculum for students whose families can afford them, while the 250 million students in government schools continue with largely unchanged pedagogy. The technology gap between private and government education, already wide, is being stretched further by AI - not because government schools are hostile to technology but because they lack the infrastructure, connectivity, and trained faculty to deploy it meaningfully.
The Cheating Crisis No One Is Talking About Clearly
The most immediate and documented problem with AI in Indian education is not dystopian - it is very mundane. Students are using AI to write assignments, essays, and project reports that are submitted as their own work. This is not a small fringe behaviour. Faculty members at several Indian universities, speaking informally, estimate that between 30% and 50% of submitted written work in undergraduate programmes now contains substantial AI-generated content. Plagiarism detection tools like Turnitin have updated to detect AI-generated text, but the cat-and-mouse dynamic between detection and evasion is well-established: students who know they are being monitored simply use AI-generated drafts as a base and modify them sufficiently to reduce AI content scores.
The more serious integrity problem involves examinations. Several documented cases have emerged of students using AI-enabled smartwatches, earpieces connected to AI question-answering systems, and in a few cases, mobile phones with AI interfaces concealed during proctored examinations. The NEET UG 2024 paper leak scandal - which had nothing to do with AI - exposed how vulnerable India's high-stakes examination infrastructure is to coordinated fraud. AI adds a new dimension to this vulnerability: unlike paper leaks, which require advance access to physical materials, AI-assisted cheating in real-time examinations requires only a connected device, which becomes harder to detect as devices miniaturise.
The Personalisation Promise: Real and Limited
Strip away both the utopian marketing of edtech companies and the dystopian fears of critics, and AI in education offers one genuinely significant benefit: the ability to personalise learning at scale. A single teacher in a class of 60 students cannot monitor whether each student is following the explanation, identify which of the 60 has a specific conceptual gap, and provide targeted practice for that gap in real time. An AI tutoring system can do all three simultaneously for every student in the class. This is not a trivial capability. Evidence from controlled studies - primarily from the United States and China, with limited Indian data - suggests that well-implemented AI tutoring can meaningfully accelerate learning outcomes, particularly for students in the middle of the ability distribution who receive less attention in a teacher-focused classroom.
The catch is the word 'well-implemented.' Most AI educational tools deployed in India have been designed for English-language learners, for academic content that maps closely to standardised curricula, and for students with reliable connectivity and devices. They perform poorly on regional language content, on subjects that require sustained writing and critical analysis rather than correct-answer identification, and on students who lack the digital literacy to navigate AI interfaces effectively. The personalisation promise, in other words, is real for the segment of Indian students for whom it works - and largely irrelevant for the majority who are outside that segment.
What the NEP Says and What Schools Are Actually Doing
The National Education Policy 2020 mentions artificial intelligence and digital technologies as important areas for curriculum integration but does not provide specific guidance on how AI tools should or should not be used in learning and assessment. The Ministry of Education has issued advisories encouraging digital literacy and caution about AI-generated content, but no binding regulation on AI use in schools or examinations has been issued at the national level. Individual states have taken varied approaches: some have issued blanket warnings against AI use in assessments, others have encouraged AI literacy programmes, and most have issued nothing at all.
The absence of a clear national policy framework is particularly significant given India's scale and diversity. In the vacuum left by policy, individual schools are making ad hoc decisions: some ban phones entirely, some require AI disclosure in assignments, some have simply chosen not to confront the issue. The result is a patchwork of incompatible local approaches that provide neither consistent student guidance nor consistent accountability standards. Students learn, quite rationally, that rules about AI use are inconsistently enforced and therefore negotiable.
The Questions Indian Education Has Not Yet Answered
The arrival of AI in Indian education forces a set of questions that the education system has not yet seriously confronted. The first is the purpose question: what exactly are we trying to achieve by educating students, and does AI help or hinder it? If the purpose of education is to transmit knowledge, AI tutors are extraordinarily efficient. If the purpose is to develop the capacity to think independently, evaluate evidence, and generate original ideas, then outsourcing cognitive work to AI may actively undermine that purpose. This is not a new debate - it predates AI - but AI has made it urgent in a way that can no longer be deferred.
The second is the equity question. AI educational tools are being deployed in a system of radical inequality. The same technology that is helping well-resourced urban students gain competitive advantages over their peers is largely absent from the government schools where the majority of Indian children are educated. If India does not make deliberate and funded efforts to close this gap, AI will not democratise education - it will entrench existing hierarchies with the added legitimacy of algorithmic neutrality.
The third is the assessment question. If students can use AI to produce written work of arbitrary quality, what does it mean to assess them on written work? India's education system, heavily reliant on terminal examinations and written assignments, has not yet produced a coherent answer to this question. Some educators argue for a return to oral and viva examination formats that are harder to AI-assist; others argue for redesigning assessments around tasks that AI cannot do well, such as collaborative problem-solving observed in real time; others argue for accepting AI as a tool and assessing students on how well they use it. None of these positions has yet been adopted as policy.
The arrival of AI in Indian classrooms is not a future event to be planned for - it is a present reality to be managed, and managed poorly or not at all. The students who have access to AI tutoring and use it intelligently will gain real advantages; those who use it to outsource their thinking will accumulate credentials without capability, with consequences they will encounter in the labour market. The students who have no access will fall further behind. The teachers who are given no guidance will make inconsistent decisions. The policymakers who defer the hard questions about assessment and equity will face a much larger crisis when they can no longer be deferred. India has never had an education technology transformation of this scale and speed. Whether it produces better-educated citizens or a more credentialed but less capable workforce depends on decisions being made right now - often by default rather than by design. That is a debate worth having before the decisions are made for us.