Introduction
India's DPI success—exemplified by the unified digital identity of Aadhaar, UPI processing nearly half of global real-time payments, and Account Aggregator enabling secure data exchange—proves that open digital networks can democratise critical services. Building upon this foundation, the ₹10,372-crore IndiaAI Mission seeks to operationalise Artificial Intelligence as India's 'DPI 2.0', converting cutting-edge intelligence into shared, accessible public utilities.
Transforming Technology into DPI: The India Stack Experience
The transition of proprietary technology into public infrastructure in India relies on three fundamental architectural principles:
- Open Protocols: Unbundling proprietary monopolies into interoperable public rails, as demonstrated by UPI for instant digital transactions.
- Consent Architecture: Guaranteeing data sovereignty and empowerment via user-centric federated models, such as the Data Empowerment and Protection Architecture (DEPA) powering Account Aggregators.
- Public Baselines: Creating state-backed foundational digital layers that lower user acquisition costs and crowd in extensive private sector innovation.
AI as the Next Frontier of Digital Public Infrastructure
AI can transcend private enterprise silos to become a scalable public good across several critical pillars:
- Compute as a Public Utility: Democratising high-performance computational infrastructure by providing subsidised access to more than 10,000 GPUs for startups, researchers, and academia under the IndiaAI Mission.
- Open Data Repositories: Establishing the India Dataset Platform (IDP) to supply curated, anonymised non-personal datasets across critical sectors like healthcare, agriculture, and climate to train indigenous models.
- Open-Source Foundation Models: Deploying shared cognitive layers such as Project Bhashini for real-time multilingual translation and BharatGen to eliminate literacy and vernacular barriers in public service delivery.
- Trustworthy Governance Rails: Embedding algorithmic transparency and safety guidelines under the Trustworthy AI Framework alongside the Digital Personal Data Protection (DPDP) Act, 2023, to mitigate demographic bias and safeguard citizen rights.
Conclusion
By treating high-capacity compute, diverse datasets, and foundational language models as public utilities, India can shift from being a mere consumer of foreign AI technologies to an architect of global digital public goods. This DPI-led model promises to upgrade public administration from reactive intervention to proactive, predictive welfare.