India has made significant strides in expanding access to education. However, a persistent gap remains between academic learning and industry requirements. While the National Education Policy laid out a strong pathway for shifting education away from rote memorisation and toward critical thinking, problem-solving, and research-driven learning, implementation has been uneven and slower than anticipated.
With AI adoption accelerating, the Indian government has initiated a major overhaul of AI education across engineering institutions, with a renewed focus on practical exposure, industry integration, faculty development, and shared infrastructure. Recently, the Union Minister for Electronics and IT, Ashwini Vaishnaw, chaired a high-level meeting with the AI Curriculum Taskforce in New Delhi to discuss reforms to align technical education with the rapidly evolving requirements of the AI industry.
The task force, working in collaboration with industry experts and NASSCOM, conducted a baseline study of existing BTech Computer Science and allied curricula across Indian institutions. While the study noted that AI-related coursework has significantly expanded in recent years, it also identified major gaps in pedagogy, infrastructure, and hands-on exposure in emerging areas such as GenAI, Machine Learning Operations (MLOps), and foundation model development.
The AI Curriculum Taskforce recommended increasing practical exposure in AI programmes from the current 25–30% to as much as 75%, alongside deeper industry integration, faculty development, and shared infrastructure. It also recommended multiple entry-exit options for students, including certificate, diploma, and advanced diploma.
For many education leaders, the move signals a shift from theoretical instruction to outcome-based learning.
From Lecture Halls to Building AI Products
Talking to AIM, Nishant Chandra, Co-founder of edtech platform Newton School of Technology, says increasing practical exposure will require institutions to rethink the entire structure of academic delivery rather than simply adding more laboratory sessions.
"Moving practical exposure in AI from 30% to 75% will separate serious institutions from cosmetic AI programmes," he says.
Chandra argues that AI education should be centred around continuous building and experimentation throughout a semester. Students should work with datasets, develop models, evaluate outputs, deploy prototypes, document decision-making processes, and understand system failures as part of the learning journey. Students must graduate with demonstrable work rather than simply honing academic credentials.
"The strongest institutions will move from lecture-first learning to project-first learning. Every semester should have real use cases, mentor reviews, practitioner feedback, and portfolio-based evaluation," he states.
Faculty Shortages are the Biggest Bottleneck
While curriculum reforms are relatively straightforward on paper, the availability of qualified AI faculty remains a major challenge.
Chandra mentions the solution lies not only in hiring more professors but also in fundamentally redesigning how AI education is delivered. "India's AI faculty shortage cannot be solved only through hiring. The demand is moving faster than traditional faculty pipelines," he mentions.
He advocates structured train-the-trainer programmes, modern teaching tools, shared learning resources, and stronger involvement from industry practitioners.
Under such a model, academic faculty would focus on theoretical foundations while industry professionals contribute real-world production experience in areas such as Generative AI, MLOps, AI agents, deployment, evaluation, and responsible AI.
Infrastructure Reality Check
While the government's vision has received broad support, questions remain about whether colleges across India possess the infrastructure needed to support AI-heavy practical learning.
Harsh Kapasi, COO of Arihant Academy, believes the direction is positive but acknowledges significant implementation challenges.
"The heightened emphasis on practical experience shows the way in which AI-based competencies are evolving as key requirements for most industries," he says.
Kapasi expects institutions to increasingly adopt project-based learning, workshops, case studies, and real-world AI applications from the first year itself. However, he notes that infrastructure disparities remain substantial.
He points to ongoing discussions around shared GPU infrastructure, faculty training, and modernisation of AI laboratories as critical enablers of the government's plans.
Will IITs Benefit the Most?
The question of equitable implementation also remains a core concern.
Kapasi believes premier institutions such as the IITs will naturally move faster because they already possess mature research ecosystems, computing infrastructure, and faculty resources.
"The IITs already have strong research ecosystems in place, as well as the infrastructure to support them, so naturally they will be able to implement this initiative much more quickly than other educational institutions in India," he explains.
However, he stresses that the long-term success of India's AI ambitions depends on whether tier-2 and tier-3 institutions can also effectively adopt AI education.
He also sees online learning platforms and digital resources playing a crucial role in democratising access to AI education.
Meanwhile, Nipun Sharma, CEO of TeamLease DA, notes that the initiative represents a broader workforce transformation strategy aligned with India's long-term economic goals.
The company notes that AI is becoming a foundational technology that will affect virtually every industry, much like digital technologies transformed workplaces over the past two decades. "The objective is not simply to create AI specialists, but to build a workforce that is comfortable working alongside AI," Sharma says.
One of the biggest risks facing universities is the speed at which AI technologies evolve.
Sharma argues that traditional curriculum cycles are often too slow to keep pace with industry developments. "The challenge is not training more people in AI, but ensuring that learning remains aligned with how AI is being used in industry," he states.
The solution lies in moving toward continuous learning models supported by industry participation, project-based learning, internships, and work-integrated experiences.
"AI education should focus less on specific tools and more on foundational capabilities such as data literacy, critical thinking, problem-solving, and the ability to adapt to new technologies," he explains.
Industry Participation Without Vendor Lock-In
As industry involvement increases, concerns have emerged around potential dependence on specific platforms or technology providers.
Chandra warns against turning academic programmes into vendor-training exercises. "The danger is not industry participation. The danger is lazy curriculum design that turns education into vendor training," he states.
"The outcome should be judgement, not vendor familiarity," he believes.
Perhaps the most far-reaching recommendation is the government's proposal to introduce AI literacy for non-STEM disciplines.
Sharma believes AI literacy will soon become as fundamental as digital literacy is today. "The impact of AI is no longer limited to technology roles. It is already influencing marketing, media, design, healthcare, finance, education, legal services, and a wide range of other professions," he states.
Rather than requiring everyone to build AI systems, future workplaces will increasingly expect professionals to understand how to use AI tools effectively, interpret outputs responsibly, and integrate them into daily workflows.
"Just as digital literacy evolved from being an advantage to a basic workplace requirement, AI literacy is likely to follow a similar path," Sharma concludes.
Your reaction