Democratising AI infrastructure: Government outlines how India plans to expand access to compute, data and models

Artificial Intelligence

• India aims to democratise AI infrastructure for wider, inclusive innovation.
• Govt boosts domestic compute capacity with IndiaAI and supercomputing initiatives.
• Digital platforms and public infrastructure to boost AI access and data sharing.

Did our AI summary help?

The Office of the Principal Scientific Adviser to the Government of India has outlined a detailed policy vision for democratising access to artificial intelligence infrastructure, positioning it as a critical priority for India’s digital and economic future. In a white paper released as part of India’s AI Policy Priorities series, the office explains how wider and more affordable access to compute power, datasets and AI models can enable innovation beyond large corporations and major technology hubs, while supporting responsible and inclusive AI adoption nationwide.

Why access to AI infrastructure matters

The white paper defines AI infrastructure as the combination of computing resources, data repositories and model ecosystems that power modern artificial intelligence systems. At present, these resources are concentrated among a limited number of global technology firms and urban centres. The government argues that this concentration restricts equitable participation and slows innovation across sectors such as healthcare, agriculture, education and public services.

By treating AI infrastructure as a shared national resource, the government aims to ensure that startups, universities, public institutions and researchers across India can participate in building and deploying AI systems. This approach aligns with India’s broader AI governance strategy, which identifies access to foundational resources as essential for both innovation and risk mitigation.

Government focus on compute and physical infrastructure

On the physical side, the government has highlighted the need to expand domestic compute capacity. Data centres, high-performance computing clusters and specialised processors such as GPUs and TPUs form the backbone of AI development. While India generates a large volume of data, its share of global data centre capacity remains limited.

To address this gap, initiatives under the IndiaAI Mission and the National Supercomputing Mission are expanding access to high-performance compute resources. Government-supported GPU clusters and supercomputers are being deployed across academic and research institutions, reducing dependence on foreign cloud providers and lowering entry barriers for AI development.

Expanding access to data and digital platforms

The government has also emphasised the role of digital infrastructure in democratising AI. Platforms such as IndiaAIKosh are designed to function as national repositories for datasets, models and tools, offering permission-based access while maintaining data ownership and privacy safeguards. Sector-specific and language-focused initiatives, including the National Language Translation Mission, aim to improve representation of Indian languages and contexts in AI systems.

State-level efforts are also playing a role, with data exchange platforms enabling controlled and federated sharing of datasets without centralised pooling. These systems allow multiple stakeholders to collaborate on AI development while preserving data sovereignty.

Digital Public Infrastructure approach to AI

A key pillar of the government’s strategy is the use of a Digital Public Infrastructure approach for AI. Rather than relying on a single centralised system, the DPI model promotes modular, interoperable layers that provide predictable and transparent access to AI resources. The government sees this approach as particularly relevant given India’s experience in building large-scale digital systems.

The white paper stresses that DPI for AI must evolve in phases, starting with directories, metadata standards and access protocols before moving to more advanced federated data and compute mechanisms. Governance, privacy protection and accountability are identified as critical requirements as access scales.

Balancing access, sustainability and adoption

While expanding access, the government has also acknowledged challenges related to energy use, sustainability and uneven sectoral adoption. Scaling AI infrastructure will increase demand for power and cooling, requiring coordination with renewable energy policies and efficient design. At the same time, targeted efforts are needed to ensure that sectors such as agriculture, healthcare and education are not left behind.