The Government of India has shared its vision for making artificial intelligence (AI) infrastructure more accessible across the country. The Office of the Principal Scientific Adviser (PSA) has released a white paper titled “Democratising Access to AI Infrastructure”, which explains how India can expand access to key AI resources such as data, computing power and digital platforms.
According to the PSA, the white paper was prepared with inputs from experts and stakeholders. It focuses on how AI infrastructure can be made affordable and available to a wider group of users, including startups, researchers, public institutions and innovators outside major cities. The idea is to ensure that AI development is not limited to a few large companies or urban centres.
The paper explains that democratising AI access means treating AI infrastructure as a shared national resource. This would allow innovators across India to build tools in local languages, develop assistive technologies and create solutions that match the country’s diverse needs. It highlights that access to computing power, datasets and AI models should be seen as essential building blocks, similar to public utilities.
The document divides AI infrastructure into two main parts: physical and digital. Physical infrastructure includes data centres, GPUs, TPUs and other specialised hardware needed to train and run AI models. While India generates nearly 20 percent of the world’s data, it currently has only about 3 percent of global data centre capacity. To address this gap, the government is working under the IndiaAI Mission to build a large GPU cluster with around 30,000 next-generation units for strategic and national use.
The white paper also notes that data centres in India are concentrated in a few cities such as Mumbai, Chennai, Bengaluru, Hyderabad and Delhi NCR. This makes it harder for smaller institutions and regional players to access advanced computing resources.
A key idea in the paper is adopting a “digital public infrastructure” or DPI approach. This means treating AI systems as digital public goods. Instead of creating one large platform, the PSA suggests building modular systems such as directories, data standards, access rules and registries. Over time, these can support secure data sharing, consent-based access and coordinated sharing of computing resources.
The white paper does not suggest new laws or policies. Instead, it presents a long-term vision to shape India’s AI infrastructure early, so that future growth remains inclusive, efficient and accessible to all.
