New plan outlines how India will democratise AI infrastructure

India is moving to rebalance access to AI infrastructure as part of a new national push to close gaps in computing power and data availability.

A white paper released in December 2025 by the Principal Scientific Adviser outlines a strategy to treat AI compute, datasets and models as Digital Public Goods, rather than resources concentrated in a handful of urban hubs.

Despite generating nearly one-fifth of the world’s data, India currently hosts only a small share of global data centre capacity. The paper outlines plans to nearly tenfold capacity expansion by 2030, alongside the rollout of national computing resources through the IndiaAI Mission.

A central pool of GPUs and TPUs is being offered at subsidised rates to researchers and startups, aiming to reduce dependence on foreign cloud providers.

Data access and sovereignty form another pillar of the roadmap. Platforms such as IndiaAIKosh and Bhashini are being developed as shared repositories, hosting thousands of datasets and models across sectors including healthcare, agriculture and Indian languages.

High-performance computing initiatives, including the AIRAWAT supercomputer, are supporting large-scale research in areas such as climate modelling and drug discovery.

The strategy also emphasises regional and state-led infrastructure, with initiatives like Telangana’s federated data exchange seeking to decentralise AI development. Sustainability requirements are also being introduced, as data centres are expected to account for an increasing share of electricity use.

Policymakers view the approach as crucial to developing a form of sovereign AI that fosters innovation beyond major technology hubs and across the broader economy.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!