New Delhi: India’s artificial intelligence strategy moved into the global spotlight this week after a pointed exchange at the World Economic Forum in Davos.
What started as a general conversation on how AI is spreading across economies turned into a firm assertion from India that it no longer belongs in any secondary category of AI nations.
On the sidelines of the Davos meetings, Union Minister for Electronics and IT Ashwini Vaishnaw responded to comments by Kristalina Georgieva of the International Monetary Fund. Vaishnaw said India is “clearly in the first group” of AI powers, and backed that claim by explaining how the country is building strength across the full AI stack.
Why India says it belongs in the first AI group
Speaking at Davos, Vaishnaw questioned how global institutions classify AI readiness. “I don’t know what the IMF criteria has been,” he said, before pointing to Stanford rankings that place India among the top countries for AI penetration, preparedness, and talent.
“Stanford places India as third in terms of AI penetration, in terms of AI preparedness, and in terms of AI talent. All three, actually, on AI talent, it is number two,” he said. “So I don’t think your classification in the second bouquet is right. It’s actually in the first.”
Georgieva had earlier said, “Actually, India would be one of them because of the bet on IT that India is making that is in the higher spectrum,” while noting that many lower income countries were “way, way behind.”
India’s five layer AI stack explained
Vaishnaw framed India’s approach around five layers of the AI architecture. His point was simple. Progress at one layer alone does not deliver results.
Application layer
This is the layer closest to people and businesses. India’s focus is on wide use, not small pilots. AI is being pushed into agriculture, healthcare, education, manufacturing, and governance. Value shows up only when adoption spreads across the population.
Model layer
Models drive intelligence behind applications. Early frontier models showed promise but needed heavy spending. Open source models reduced costs. India is building models that work for Indian languages, sectors, and rules. Sovereign models matter for data safety and cultural fit.
Chip and compute layer
AI needs compute like GPUs and specialised chips. Access matters. Under national missions, more than 38,000 GPUs are being offered at about one third of the global average cost. Work is also underway on semiconductor fabs and ATMP units.
Data centre layer
Data centres form the digital backbone. India has already attracted close to $70 billion in investments from firms like Google, Microsoft, and Amazon. Efforts are ongoing to improve cooling, water use, and energy efficiency.
Energy layer
AI needs steady power. Renewables help but are not constant. Nuclear energy is being positioned as stable baseload power under the SHANTI Act, using small modular and micro reactors with public private partnerships.
Why this debate matters now
Vaishnaw stressed that results come from use, not size alone. “ROI doesn’t come from creating a very large model,” he said, adding that most use cases work with models in the 20 to 50 billion parameter range.
With India set to host an AI summit next month, the Davos exchange signals intent. India wants to define its AI path through infrastructure, talent, and real world use, and not wait for outside labels to decide its place.
