For AI, India can build on the Aadhaar-UPI model

For India, AI is a leapfrog moment. We already have digital identity, payments, and platforms operating at scale; we can move directly to the intelligence layer, combining our data and green energy to build the world’s most inclusive AI economy. To realise this, three priorities are essential.

First, AI systems should be subject to our laws and strategic priorities. The most capable AI systems today are foundation models trained on massive datasets compiled through the interactions of billions of users. For India, the critical issue is not ownership but jurisdiction. If the systems shaping our education, healthcare, or commerce operate abroad, we lose oversight of data use, algorithmic behaviour, and accountability. Foundation models that serve Indian users must therefore be trained, hosted, and regulated within the country.

This is not only a matter of digital sovereignty but also of national security. In the years ahead, AI systems will manage public infrastructure, guide administrative decisions, and process sensitive data across defence, energy, and finance. India’s approach must be distinct from both the corporate ownership of the West and the state centralisation of China. Just as we built our own identity and payment systems, we now need to develop our own intelligence systems.

Second, we must build out our own green AI infrastructure. AI systems depend on vast computing power. Most of it today is concentrated in the US and China. India must rapidly expand its computing capacity. Developing this infrastructure is a massive green industrial opportunity. Data centres powered by renewable energy can anchor new investment in solar, wind, and green hydrogen. They can stimulate innovation in semiconductors, power electronics, and advanced cooling systems. With the right policy and financing, India can build one of the world’s best low-carbon AI infrastructures.

It is a leap to the green frontier where we combine technology competitiveness with long-term sustainability.

Third, every Indian should have a multilingual agent that they control, and that will assist them in this brave new AI world. A farmer could use his agent for weather and crop advice, a student for language learning, a patient for managing medical records. These assistants must be private, transparent, and accountable, with data stored securely under user control. Creating these agents will require open standards, privacy safeguards, and collaboration between government, academia, and startups. The same principles that made Aadhaar and UPI reliable and inclusive can guide this next layer of DPI.

AI will change how work is organised and where value is created. Some existing service roles will shrink, but new opportunities will emerge in data infrastructure, model development, and domain-specific applications. India’s task is to move early, invest in capability, and capture these higher-value activities. Note that China used the clean-energy revolution to dominate global supply chains for solar, batteries, and electric mobility. We can now use our low-carbon AI systems to build the world’s best service delivery capabilities.

If large AI systems operate under Indian law, if computing power grows through green infrastructure, and if citizens can access AI securely and affordably, we can leapfrog directly to the green frontier.

The writer is president, Everstone Group and visiting professor in practice at the London School of Economics. He is a former Union minister and Lok Sabha MP. Views are personal