A background note can be accessed here: NITI Aayog’s roadmap for the next phase of India’s DPI
The DPI@2047 roadmap shifts from centrally driven rollouts to a State- and district-led, problem-first execution model with shorter implementation cycles. How does this decentralisation affect the integrity of India’s interoperable digital stack?
Decentralised execution is a necessary progression as DPI expands beyond foundational systems such as Aadhaar and UPI into sectors such as agriculture, MSMEs, health, and education. These sectors are shaped by highly localised administrative realities, service gaps, and institutional capacities that cannot be effectively addressed through a uniform central design architecture.
The challenge lies in preserving interoperability while enabling localised experimentation. Fragmentation may emerge not only from incompatible technical standards, but also from legislative divergence, linguistic and semantic inconsistencies, uneven institutional capacity, and limited incentives for states to exchange data or services. India’s experience with land-record digitisation already demonstrates how some states become institutional frontrunners while others struggle to operationalise reforms at scale.
The policy priority, therefore, is to build a federated architecture anchored in nationally coordinated governance principles. Interoperability must be embedded during system design rather than treated as a downstream integration exercise. Common frameworks for identity, consent, APIs, registries, and data exchange will need continuous testing before federated deployment. A national regulatory and standards-setting mechanism can help maintain coherence while still allowing state-led innovation and context-specific implementation.
The framework emphasises “unlocking data” through verifiable credentials and expanding AI-driven services to enable productivity and inclusion. To what extent can this expansion of data use sustain trust and accountability in India’s DPI ecosystem?
DPI 2.0 positions interoperable data exchange as a core public infrastructure layer. The objective is to connect siloed public and private datasets through verifiable credentials, anonymised exchanges, and consent-based architectures that can support inclusion and productivity. A farmer combining land records, crop histories, and mandi transaction data to access credit, or urban planners using aggregated mobility data for transport planning, illustrates the developmental potential of such systems.
However, scaling data use also expands governance responsibilities. Risks associated with data leaks, unauthorised appropriation, opaque AI-driven decision-making, and excessive concentration of control over data infrastructures could weaken institutional trust if safeguards remain uneven across states and sectors. Trust in DPI will increasingly depend on whether governance systems evolve alongside technological capacity.
This requires robust institutional infrastructure at the state and district levels: granular access controls, audit trails, consent receipts, third-party security audits, grievance redressal systems, and clearly assigned liability frameworks. Long-term sustainability is equally important. If DPI ecosystems struggle to recover infrastructure costs, commercial incentives may gradually shift toward monetisation of data assets and function creep. Stable public financing models and strong public data institutions will therefore be central to maintaining accountability within an expanding DPI ecosystem.
DPI 2.0 marks a shift from access-led inclusion toward livelihoods and Total Factor Productivity gains, particularly for lower- and middle-income segments. How should policymakers navigate potential trade-offs between efficiency gains and equitable access?
DPI 2.0 seeks to improve productivity by enabling farmers, MSMEs, and citizens to access credit, markets, advisories, and public services through interoperable digital systems. The geographically distributed state-champion approach is strategically valuable because it allows sectoral and regional experimentation before wider adoption. Yet productivity-oriented systems can also reproduce exclusion if underlying datasets fail to capture the realities of vulnerable populations.
AgriStack offers a useful illustration. If farmer databases are structured primarily around land ownership records, tenant farmers, sharecroppers, women farmers, and landless labourers may remain outside the service ecosystem. Comparable exclusion risks may emerge across health, MSME financing, and welfare delivery. Digital access gaps also persist across gender, age, and income groups despite rising smartphone penetration.
Policy design must therefore combine technological efficiency with layered inclusion strategies. Localised interfaces, including vernacular and voice-based systems, can improve accessibility, but they may not fully address exclusion among populations with limited digital capability or internet access. A durable transition strategy may require phygital institutional models that combine digital systems with assisted offline service delivery. Bangladesh’s one-stop service centres operated by local women entrepreneurs demonstrate how intermediary institutions can help citizens access DPI-linked services while gradually integrating them into the digital ecosystem itself.