– Jayant Rastogi, Global CEO, Magic Bus India Foundation

In a small training centre on the outskirts of Ranchi, a group of young people gather around a single smartphone. The trainer asks, “How would you explain teamwork in a job interview?” One student speaks into the phone. An AI tool responds with suggestions, refining the answer in simple English. The group listens, repeats, laughs at mistakes, and tries again.

There is no computer lab, no broadband, and no personal devices. And yet, something is working.

This is the kind of classroom India’s evolving AI-skilling efforts are beginning to design for.

The Design Assumption We Need to Revisit

Across conferences, policy discussions, and product launches, AI is often positioned as a great equaliser for India’s youth. The promise is compelling. AI can personalise learning, break language barriers, and expand access to knowledge.

There is truth in this, and we are already seeing early signs of what is possible. At the same time, many of these efforts are built on an important assumption: that the learner has access to a personal device, most often a smartphone with reliable connectivity, and sometimes even a laptop.

On paper, this seems reasonable. A large proportion of youth in Tier II and Tier III India live in households that own a smartphone, and many can access one during the day. Access is an important first step, but translating it into meaningful learning requires additional support and thoughtful design.

This distinction is increasingly gaining attention as the ecosystem evolves. A shared household device is rarely a personal learning tool. A phone belonging to a parent or sibling may not be consistently available, especially for a young woman trying to practise interview responses or build digital confidence.

The gap between access and meaningful use is wide. When AI tools are designed with these realities in mind, they have the potential to reach those who stand to benefit the most.

What Frugal Design Really Means

It is important to understand that “frugal innovation,” often interpreted as the development of low-cost solutions, means something more intentional in the context of AI for skilling. It is about designing for constraints as the primary reality, not as an exception.

This means building tools that operate over low-bandwidth connections and remain usable even when connectivity drops. It also means designing for users who are more comfortable speaking than typing, especially in communities where oral communication is the norm. Language support must go beyond translation to reflect how people actually communicate, whether in Nagpuri, Gondi, Santali, or other regional contexts.

It also requires acknowledging that youth from rural and low-income backgrounds are significantly less likely to be digitally skilled than their urban peers. If AI tools are designed for a highly connected, digitally fluent user and then adapted for everyone else, they may not fully bridge existing gaps. This highlights the importance of inclusive design from the outset.

Many product teams building these solutions are based in urban centres, and their assumptions about how learners engage with technology are shaped by their own experiences. Addressing this requires a structural shift towards genuine co-design with the communities and facilitators who are closest to the learner.

The Role of the Facilitator

In many conversations around AI and education, the human facilitator is often seen as a constraint, an intermediary that technology can eventually replace.

Our experience suggests the opposite. In underserved communities, the facilitator is often the most critical influence in a young person’s learning journey. They understand context in a way no system can. They know which learner is struggling and why, who is at risk of dropping out, and how to build confidence in those who hesitate to speak.

For the young people we work with, facilitators do more than impart knowledge. They nurture confidence, social communication, teamwork, and resilience — the life and employability skills that employers repeatedly say they value most.

Technology cannot replace this layer. It can, however, strengthen it.

AI has the potential to give facilitators better tools, more timely insights, and additional ways to support learners. It can enable more personalised feedback and create opportunities for practice that were not previously possible. But its role should be to augment human interaction, not replace it.

The most effective models will be those where technology and human support work together, each enhancing the other.

Beyond Skills: Building Agency

According to the Mercer Mettl India Graduate Skill Index 2025, India’s graduate employability rate stood at 42.6% in 2024. While this highlights the urgency of skilling, it also points to a deeper issue.

Employers consistently emphasise the importance of life skills such as communication, adaptability, problem-solving, and the ability to work with others. These are not skills that can be acquired passively. The World Economic Forum’s Future of Jobs Report highlights analytical thinking, resilience, and communication as some of the most critical skills for the future workforce.

These are life skills that develop through guidance, practice, feedback, and reflection. AI can accelerate the process through simulated conversations, scenario-based prompts, and instant feedback on spoken English. But learners need enough agency and confidence to engage critically with the tool, question its outputs, and apply their own judgment to the interaction.

Building that agency is the real work. AI supports that work.

Designing for the Masses

India is making significant investments in AI, skilling, and digital infrastructure, not just as an economic strategy, but with a broader ambition to create inclusive growth. With initiatives such as the IndiaAI Mission, backed by substantial public investment, the country is well-positioned to build an AI ecosystem that is both innovative and inclusive.

At the same time, continued focus on on-ground realities will help ensure that AI strengthens its role as a powerful equaliser across diverse communities.

This makes it important to ask a simple but critical question before any solution scales: Who is it truly designed for? Does it work on the devices learners actually have access to? Can it function reliably when connectivity is inconsistent? Does it reflect the way people communicate in their own languages? And does it strengthen the role of the facilitator, who remains central to the learning experience?

The students in that classroom in Ranchi are the majority. Designing with their realities at the centre creates an opportunity to ensure that technology meaningfully expands access and opportunity.

The future of AI should not belong only to those with the best devices or the fastest internet.

It should belong to everyone.

Also Read: Why simulation-based training is becoming essential in India’s medical colleges