For the better part of a century, India’s education system has trained humans to behave like machines—to calculate, memorise and follow logic—and perfected that craft just as machines began doing all three better than we could.
We still mass-produce the very skills artificial intelligence (AI) has made abundant and cheap, while the capacities AI cannot mimic—judgement, curiosity, the ability to ask the right question—remain an afterthought.
The old promise, that a skill and a degree guaranteed a stable career, is being dismantled by technology itself. India’s education policy must learn how to keep learning.
The journey so far: Our policy has always been reactive to its moment. In 1968, a young nation used education to forge a national identity and basic literacy. In 1986, Operation Blackboard carried schooling into rural India while pivoting to computer literacy that fuelled a global information technology (IT) boom.
The National Education Policy (NEP) 2020 was the boldest reset yet: its 5+3+3+4 structure, a move beyond rigid science-commerce-arts stream silos, multidisciplinary learning and vocational training in the mainstream.
It was the right policy for 2020. But a once-in-a-generation policy cannot keep pace with a technology that reinvents itself every few months. Knowledge has become a commodity; intelligence is now the differentiator.
The opportunity and risk: As one of the world’s youngest nations, India can harness a vast demographic dividend, but only if we nurture it before it turns into a liability. India produces nearly five million graduates a year. Yet, according to the State of Working India 2026 report, only about 40-45% are employable.
For some 1.5 million engineers, Indian industry has only about 200,000 high-end roles. The rest drift into call centres or low-value gig work, and other streams fare no better. Employers find fresh graduates so inadequate that they retrain them for 6-12 months.
With AI rapidly automating routine roles, we are racing to prevent a surge of degree-carrying, unemployable youth—and to turn that pipeline into a workforce of employable assets.
A fluid curriculum: NEP 2020 laid excellent groundwork. We now need a structural recalibration towards a fluid curriculum where the learning blocks are modular and continuously updated, not frozen for five years. Primary school should be an incubator of curiosity and collaboration, not a store of facts.
Secondary school should shift from learning by rote to reasoning, analytical judgment and prompt engineering—and close the missing link of vocational exposure: let students in grades 6-9 shadow a local electrician, baker or coder, and those in grades 11–12 take 8-10 week internships that earn the same credits as regular courses in physics or mathematics.
Credit parity is what makes students, parents and academia take skilling seriously.
At the university level, single stream specialization has become a pathway to obsolescence and is fast becoming a ticket to a dead end. The future is interdisciplinary—agricultural science with AI-driven climate modelling, or public policy with cybersecurity.
Every degree should offer vocational majors with equal credit and path flexibility so that a student can step out after two years with a vocational certificate, work for some time, and then return to finish the degree without any penalty.
Widen out learning: Our tier-2, tier-3 and tier-4 towns will drive the next decade. Rather than ask every school to build a computer lab, 3D printer or a solar lab, set up a well-equipped Industrial Training Institute in each district as a central training hub.
Let geography shape the syllabus: One District One Product has given towns a new identity—Kancheepuram for silk, Moradabad for brass, Tirupur for textiles—so a school in a textile hub should teach textile design and loom maintenance, not generic mechanical engineering.
Local micro, small and medium enterprises (MSMEs) can become apprenticeship avenues for local hubs, drawn in by tax breaks or subsidies, turning students into micro-entrepreneurs who build where they belong.
Make peace with AI: The fear that AI will take jobs is real but incomplete. AI will not replace the doctor, lawyer or sales executive. But the one who uses AI could replace the one who does not. It amplifies human potential rather than substituting for it.
AI literacy must, therefore, become as foundational as language or mathematics from the early years, and domain-specific across every stream—a commerce student using AI for tax and reporting, an engineer for predictive maintenance.
At the degree level, a hands-on apprenticeship deploying a real AI solution should be a prerequisite for a student to graduate.
Eternal learning: Learning has to be non-stop. That mindset must travel from our boardrooms to our classrooms and policy corridors. The next phase must aim to produce future-ready learners who can adapt to employers’ needs as they evolve.
The Union budget for 2026-27 has sketched an ambitious map, from university townships in industrial corridors to AI-enabled curricula. But as we pivot towards a Viksit Bharat, we must strengthen the weakest link: execution.
If we do not close the gap between policy intent and classroom reality, we will leave this generation stranded in the analogue past while the world moves to a cognitive-first future. Let us not merely prepare for that future—let us architect it.
The author is chief operating officer, IndiaMART InterMESH Ltd.