NEW DELHI: Former NITI Aayog CEO Amitabh Kant said India is contributing more data to AI development than the United States. The G20 Sherpa has also urged the country to build its own AI models to benefit from this contribution.
Speaking at the India AI Summit 2026, Kant said: “If you look at Open AI, ChatGPT, we are providing more data than the United States of America, 33% more data than what the United States of America do. These large language models are getting better and better on the basis of data from the Global South. It is essential that this contribution translates into benefits for these regions.”
He has also warned that without a framework for Digital Public Infrastructure (DPI), the world risks creating a deeply unequal global society. He said AI’s evolution must be guided by three pillars: accessibility, affordability, and accountability.
Kant highlighted a significant disparity in how large language models (LLMs) are trained, noting that the Global South is the engine room of AI development. He argued that for AI to be inclusive, it must move beyond English-centric models and become natively multilingual to serve diverse populations.
Why Amitabh Kant wants AI to follow blueprint of India’s DPI
Drawing parallels with India’s success in financial inclusion, Kant suggested that AI should follow the blueprint of the country’s Digital Public Infrastructure (DPI), which enabled India to achieve decades of development in seven years. DPI is an open, interoperable digital framework for identity, payments, and data exchange that helps governments deliver secure and accessible public services at scale.
Agencies