Getting AI right, the Indian way

Adoption must be driven by context, judgment, and the institutional strength for local problem-solving.

As India prepares to host the mega AI Impact Summit, the prevailing mood is one of excitement. That is understandable. Artificial Intelligence promises scale, speed, and the possibility of leapfrogging long-standing constraints. Few technologies in recent memory have generated this level of global attention. Yet moments like these also deserve restraint.

Not because AI is overhyped, but because history suggests that when enthusiasm runs ahead of judgment, outcomes suffer. I find myself reflective rather than euphoric, not out of scepticism, but out of lived experience. I felt the same way in the early days of the internet, and later during the outsourcing boom. In both moments, global experts arrived with confidence and compelling narratives. Predictions were bold. Solutions were neatly packaged and sold. The future, it seemed, had already arrived. It hadn’t.

In hindsight, the problem was not the technology. It was borrowed thinking. Context was ignored, local constraints were underestimated, and judgment was outsourced to those who understood the tools but not the terrain. Many promised outcomes took far longer to materialise, and some never did. That risk feels very present again.

We will see a familiar pattern play out. Prominent global leaders will arrive with fanfare, offering confident predictions about how the future might unfold. Their demonstrations will be impressive and their narratives persuasive. Yet the one certainty I feel most strongly is not about how quickly societies will change, but about how prominently India will feature in their growth plans. When certainty is this absolute, it often reflects clarity about commercial opportunity more than clarity about real-world outcomes. This is where judgment matters.

When countries and organisations listen too closely to experts selling certainty and subscriptions, a predictable sequence follows. Automation begins before the work itself is deeply understood. Tools are deployed before roles are redesigned. Metrics of “impact” are chased before anyone pauses to ask what kind of work should exist in the first place.

Technology vendors are rewarded for speed and scale. Nations, however, pay the price when judgment is replaced by imitation. AI amplifies this risk because it scales not just efficiency, but assumptions. Clear thinking compounds. Shallow thinking spreads just as fast.

India’s opportunity with AI does not lie in becoming the largest consumer of tools or the fastest adopter of global frameworks. Tools will become commodities. Certifications will multiply. Subscription models will proliferate. None of this guarantees a meaningful impact. What matters is thinking. India does not need more gyaan. It needs its own thinking.

Over the last few years, I have seen what thoughtful AI impact can look like, particularly in rural education. For the first time, data is emerging directly from classroom teaching rather than relying only on lagging indicators or periodic surveys. That visibility changed behaviour. Education officers are now nudged to encourage schools where teaching practices are working well and, equally importantly, to intervene early where schools are falling behind. The technology itself is not the breakthrough. The breakthrough is the combination of data with human judgment, local context, and timely action. Used this way, quietly and purposefully, AI is delivering impact not just measurable, but significantly meaningful.

Beyond the summits

This is why the shift from treating AI as a demi-god to focusing on AI impact, as this summit is attempting to do, is welcome. Impact, however, does not travel well on its own. It has to be created where problems are real, constraints are non-negotiable, and trade-offs are unavoidable. The most important work of AI, therefore, is not technical. It is cultural.

It forces organisations and governments to confront uncomfortable questions. Who decides when algorithms are confident but wrong? Where does accountability sit when judgment is partially automated? What must remain deeply human as machines grow more capable? These questions are not resolved at summits. They are answered over time, in classrooms, offices, factories, and public systems, long after the spotlight has moved on.

India should be ambitious about AI. But ambition without judgment is fragile. The real test will not be how many pilots are launched or tools deployed, but whether we build the institutional muscle to learn, adapt, and think independently. The internet accelerated Indian IT because we reimagined its use rather than replicating existing models. That instinct to adapt instead of imitate made all the difference. AI will require the same approach: patience over speed, discernment over adoption, and confidence in local problem-solving rather than borrowed certainty.

That is not pessimism. That is lived experience.

(The writer is Founder-Chairman, Sampark Foundation)