India should rethink how it builds and governs artificial intelligence by treating core AI resources as shared public infrastructure rather than concentrating solely on ownership of large-scale hardware systems, according to a new white paper released by the Office of the Principal Scientific Adviser to the Government of India. The paper argues that access to essential AI inputs—such as computing power, datasets, and models—has emerged as a “critical determinant of innovation, competitiveness, and governance in the digital economy,” even as these capabilities remain heavily concentrated among a small number of global technology firms and major urban centres.
The document calls for a shift in national strategy, emphasising broad access over exclusive control. It highlights the risk that limited availability of AI infrastructure could deepen existing digital divides, slowing innovation outside well-funded corporations and metropolitan regions. Instead, the paper recommends making AI resources widely accessible through Digital Public Goods (DPGs) and Digital Public Infrastructure (DPI), drawing inspiration from India’s own large-scale platforms such as Aadhaar and the Unified Payments Interface (UPI).
Explaining this approach, the white paper states, “Democratising access to AI infrastructure means making the AI infrastructure – compute, datasets and model ecosystem available and affordable, such that it reaches a wide set of users.” To achieve this, it stresses the importance of governance frameworks that prioritise equity, affordability, and broad reach, ensuring that startups, researchers, public institutions, and smaller enterprises can all benefit from AI capabilities.
A central theme of the paper is the idea of positioning AI as a shared utility. Treating AI resources as DPGs, it explains, would involve recognising them as “shared public utilities, rather than proprietary assets.” This could be implemented through mechanisms such as open and trusted data repositories, subsidised or shared access to compute resources, and publicly supported model hubs that lower the barrier to entry for AI development.
The paper also revisits the concept of an AI commons—an ecosystem of curated, interoperable datasets designed for public-interest applications. Such an approach would focus on datasets relevant to India’s linguistic diversity, sector-specific needs, and social priorities, enabling more inclusive and locally grounded AI solutions.
By framing AI infrastructure as a public good, the white paper positions India to build a more resilient and inclusive digital economy. The proposal signals a move away from purely market-driven access toward a model that balances innovation with public interest, aiming to ensure that the benefits of AI extend beyond a narrow set of players to society at large.
