India’s Strategic Shift: How Digital Public Infrastructure is Set to Democratize the AI Revolution

The global race for Artificial Intelligence dominance is often seen as a battle of giants, with massive corporations holding the keys to the most critical resources: computing power and data. However, India is charting a different course. The Indian government has recently signaled a major shift in its AI policy, moving toward a model that treats AI infrastructure not as a proprietary tool for a few, but as a shared public good. This vision was detailed in a comprehensive white paper published on December 29, 2025, by the Office of the Principal Scientific Adviser to the Government of India.

At the heart of this strategy is the concept of Digital Public Infrastructure (DPI). By applying the same principles that revolutionized digital payments and identity in India, the government aims to build shared building blocks that allow innovators of all sizes to participate in the AI age. This approach is designed to prevent the concentration of AI resources among a handful of global firms, ensuring that the benefits of the AI revolution are distributed more equitably across the nation.

The Blueprint: Telangana Data Exchange (TGDeX)

India is not just talking about this theory; it is already putting it into practice. The Telangana Data Exchange (TGDeX) serves as the country’s first real-world example of DPI for AI. Developed by the state government of Telangana, this platform acts as a bridge between diverse sectors. It integrates datasets from government agencies, academic institutions, and the private sector into a single, cohesive platform.

The ambitions for TGDeX are significant. The goal is to create 2,000 AI-ready datasets over a five-year period, spanning from 2025 to 2030. This initiative provides a blueprint for how data can be treated as a shared resource, enabling multiple stakeholders to collaborate on AI development without the traditional friction of data silos.

Maintaining Sovereignty and Privacy

One of the most critical aspects highlighted in the white paper, titled “Democratising access to AI infrastructure,” is the focus on data sovereignty. The proposed DPI model allows for secure and privacy-compliant sharing of datasets. Crucially, this system enables collaboration without requiring the movement of raw data. By keeping the data where it resides but allowing AI models to learn from it, India is addressing one of the biggest hurdles in modern tech: how to innovate while protecting sensitive information and national interests.

The paper is the result of extensive consultations, policy engagements, and expert reviews within the AI ecosystem. It serves as a foundational document intended to shape the governance landscape of India’s AI future. For India, democratizing access is a policy priority, ensuring that a startup in a small city has the same technical opportunities as a tech titan in a major hub.

A Modular Approach to Technical Architecture

Building a national AI stack is a complex task, and the subcommittee behind the paper suggests a phased, modular approach. The goal is to identify technical pathways that connect the foundational layers of AI, specifically compute power and data. By developing a shared technical architecture, the government hopes to unify these layers, reducing fragmentation and making the use of high-performance computing more seamless for researchers and developers.

The rollout is envisioned in stages. Initially, the focus will be on “lighter weight elements.” This includes establishing directories, metadata standards, registries, and access protocols. These are the basic rules of the road that allow different systems to talk to one another. As the infrastructure matures, more “advanced elements” will be introduced, such as federated data access systems, consent-based data flows, and sophisticated coordinated compute-exchange mechanisms.

Breaking Down Barriers for Startups and Researchers

Perhaps the most significant value of the DPI approach lies in its ability to lower entry barriers. Currently, the cost of high-end computing and the difficulty of accessing high-quality datasets are prohibitive for smaller firms and research institutions. By creating predictable, transparent, and interoperable access pathways, India aims to level the playing field.

The white paper emphasizes the importance of reusable, open layers—often referred to as digital public goods. These include open data repositories, subsidized compute clouds, and open-source model hubs. By providing these resources, the government is positioning DPI as a complementary force to other interventions, such as direct investments in infrastructure expansion and capacity building.

Ultimately, India’s move toward a DPI-led AI strategy is about more than just technology; it is about creating a fair and competitive ecosystem. By treating AI infrastructure as a public resource, India is ensuring that its journey into the AI age is inclusive, transparent, and driven by innovation from every corner of the country.