Published February 2, 2026 | Version 1.0
Report Open

DPI-AI Framework: Building AI-Ready Nations through Digital Public Infrastructure

  • 1. ROR icon International Institute of Information Technology Bangalore
  • 2. Center for Digital Public Infrastructure

Description

As governments worldwide confront rapid advances in artificial intelligence (AI) alongside persistent fragmentation in public digital systems, a central challenge has emerged: how to integrate AI into the public sector in ways that are interoperable, governable, and aligned with public purpose. This paper introduces the DPI–AI Framework, a conceptual and architectural approach for embedding AI capabilities into Digital Public Infrastructure (DPI) rather than treating AI as a standalone or vertically integrated solution.

 

The framework positions AI as an external, modular, and interoperable layer that connects to foundational DPI rails—such as digital identity, data exchange, and payments—through shared standards, governance mechanisms, and safeguards. It introduces three core building blocks: AI Blocks, which encapsulate discrete AI capabilities as callable, auditable functions; DPI Workflows, which orchestrate AI Blocks with policy rules, data access, and human oversight; and Public Agents, which act as AI-enabled interfaces for citizens and public servants while remaining accountable to institutional authority.

 

Rather than prescribing specific models or technologies, the DPI–AI Framework provides a shared mental model for reasoning about AI adoption in government. It emphasizes modularity, reuse, inclusion, sovereignty, and transparency, enabling governments to introduce intelligence incrementally, modernize legacy systems through a “+1” integration approach, and avoid vendor lock-in. The paper also explores how existing Digital Public Goods can evolve into AI Blocks, how safeguards can become callable infrastructure, and how workflow-centric design enables auditable, rights-based service delivery.

 

By reframing AI as infrastructure-compatible and publicly governed, this paper argues that the value of AI in the public sector lies not in autonomy or scale alone, but in its integration with DPI to strengthen state capacity, improve service delivery, and preserve democratic accountability.

Files

Vision_Paper_-DPI-AI-Framework_2026_CDPI.pdf

Files (2.6 MB)

Name Size Download all
md5:185c03d24b1480c203047052c856a3e2
2.6 MB Preview Download

Additional details

Dates

Available
2026-02-02