CDPI Vision Paper · 2026

Building AI-Ready Nations through Digital Public Infrastructure

AI does not require a new framework — it requires the right integration with existing DPI foundations. A practical approach to embedding modular, governable intelligence into public service delivery.

3Core Elements
5Use Case Archetypes
9Implementation Steps
6+7Principles & Patterns
👥
Public Agents
AI-enabled citizen interfaces
⚙️
DPI Workflows
Orchestration & governance layer
🧱
AI Blocks
translate() · verify() · classify()
🏛️
DPI & Legacy Systems
Identity · registries · payments
💾
Data & Hardware
Sovereign cloud · on-premise
About this paper

The value of AI emerges not from its autonomy
but from how it integrates with DPI

This paper presents the DPI-AI Framework — a practical way to think about how Artificial Intelligence can be integrated into public digital systems through Digital Public Infrastructure. Written at a moment when AI capabilities are advancing rapidly while many governments still grapple with fragmented systems and legacy architectures.

Rather than proposing AI as a standalone transformation, the framework explores how existing DPI foundations can provide structure and coherence for AI in the public sector. It deliberately avoids prescribing specific technologies or models. Instead, it offers a shared mental model.

This framework does not assume AI should be prioritised ahead of foundational DPI work. In many contexts, the immediate focus remains on clean registries, reliable APIs, and minimal institutional capacity.

"Digital sovereignty is not achieved through isolation — but through capability. It depends on building the competence to design, reuse, maintain, and evolve shared public systems independently."

The paper is intended for digital leaders in government, development partners, donors, and practitioners who are shaping national and global digital agendas.

At its core, government is a system of:

📋
Registries
Records that are created, updated, or read — identity, land, benefits, credentials
⚖️
Rules
Laws and policies that are executed — eligibility, consent, approval logic

AI Blocks can read and write registries. DPI Workflows encode and execute rules. Public Agents interpret citizen intent and trigger the right combination of both.

How to cite
CDPI Abadie, D. (2026). DPI-AI Framework: Building AI-Ready Nations through Digital Public Infrastructure. https://digitalpublicinfrastructure.ai
The DPI-AI Framework

Three interrelated elements

Together they describe how intelligence can be embedded into public service delivery while preserving interoperability, accountability, inclusion, and sovereignty.

01
AI Blocks

Modular AI Capability

Callable functions that encapsulate a single AI capability — the ingredients of any AI-enabled service. Minimalist, composable, reusable, and governable.

Foundational Sector-Specific API-invoked Replaceable
Foundational: translation, speech-to-text, OCR, summarisation
Sector-specific: eligibility verify, identity verify, clinical decision support
02
DPI Workflows

The Orchestration Layer

The recipe that coordinates AI Blocks with DPI systems, policy rules, data flows, and human oversight. The workflow is the locus of control — AI operates within it, never outside it.

Identity verify Data exchange AI invocation Human review
Defined as YAML/code. Can be shared as reusable templates across governments. Embedded safeguards ensure no AI output directly modifies records without validation.
03
Public Agents

AI-Enabled Interfaces

Constrained interfaces — not decision-makers — that interact with citizens and public servants. They activate DPI Workflows and interpret citizen intent while remaining accountable extensions of government.

Multi-channel Local languages Human escalation Consent-bound
Operates on WhatsApp, voice, USSD, web. Can only activate approved workflows. Always offers human recourse. Not autonomous — accountable by design.
Architecture

6 Principles · 7 Patterns

The principles explain why the framework is designed this way. The patterns show how to apply it in practice.

🔗

Interoperability First

AI Blocks connect through open standards — not proprietary integrations. Any block can be swapped without rebuilding the workflow.

🏛️

DPI as the Source of Truth

AI outputs inform decisions; registries remain authoritative. No AI block directly modifies authoritative records without a governance step.

🧩

Minimal Footprint

Each AI Block does one thing well. Complexity is managed at the workflow level, not embedded in blocks.

👁️

Human Oversight by Default

Any low-confidence output, exception, or safeguard violation routes to a human. Automation is earned, not assumed.

🔒

Sovereignty & Replaceability

Model provenance, hosting options (sovereign cloud / on-prem / external API), and replaceability conditions are declared in every block spec.

📊

Govern Data, Not Just Models

Data provenance, consent, and stewardship are foundational. A government cannot train a trustworthy model on ungoverned data.

The +1 Approach: Governments don't need to replace legacy systems to become AI-ready. Layer new capabilities alongside what exists — wrapping legacy systems with interoperable APIs and DPI blocks. Every registry cleaned, every API published, every governance process formalised is a direct enabler of AI-ready public infrastructure.

Use Cases

Making the vision tangible

Five sector archetypes showing how AI Blocks, DPI Workflows, and Public Agents combine to deliver real services.

Social Services & Benefits — Service Journey
1
Citizen authenticates through digital identity system (e.g. MOSIP)
2
Public Agent collects benefit claim via WhatsApp or USSD in local language
3
AI Block eligibility_verify() checks against social registry and policy rules
4
AI Block document_extract() processes supporting documents (OCR)
5
Low confidence or exceptions → escalate to human caseworker
6
Approved claim triggers payment via DPI payments rail (Mojaloop/UPI)
7
Citizen receives confirmation with case ID and human recourse instructions
AI Blocks Used
eligibility_verify()document_extract()identity_verify()translate()
DPI Dependencies
Digital IdentitySocial RegistryPayments Rail
Channels
WhatsAppUSSDWeb Portal
Key safeguard: No payment is triggered without identity verification and eligibility confirmation passing through the DPI Workflow. All exceptions route to human review.
Civil Registration — Birth Certificate Issuance
1
Hospital staff submits birth notification via web form or mobile app
2
AI Block ocr_extract() processes physical birth notification forms
3
AI Block identity_verify() cross-checks parents against civil registry
4
Duplicate detection AI Block checks for existing entries
5
Human registrar reviews and approves — registry updated
6
Digital birth certificate generated and issued to parents
7
Child automatically enrolled in health & education registries
AI Blocks Used
ocr_extract()identity_verify()duplicate_detect()
DPI Dependencies
Civil Registry (OpenCRVS)Digital IDHealth Registry
Channels
Hospital SystemMobile App
Key benefit: Reduces duplicate entries and manual data entry errors. The AI contributes extraction and risk assessment; the human registrar retains decision authority over the authoritative record.
Livelihoods — Farmer Benefit Application via WhatsApp
1
Smallholder farmer sends WhatsApp voice note in local language
2
AI Block speech_to_text() transcribes in local language
3
AI Block translate() normalises to administrative language
4
AI Block structured_extract() pulls name, location, crop type, land size
5
DPI Workflow verifies identity and checks agricultural registry eligibility
6
OpenFn enrolls farmer in benefits registry and submits application
7
Farmer receives WhatsApp confirmation with case ID; ambiguity → human caseworker
AI Blocks Used
speech_to_text()translate()structured_extract()eligibility_verify()
DPI Dependencies
Digital IDAgricultural RegistryPayments (Mojaloop)
Channels
WhatsApp VoiceUSSD fallbackSMS
Key insight: Voice-first design serves low-literacy populations. Channel fallback (WhatsApp → USSD → SMS) ensures low-connectivity resilience. Local language coverage is a governance priority, not an afterthought.
Education — Personalised Learning Support
1
Student authenticates using student ID (linked to education registry)
2
Public Agent (tutoring interface) assesses current learning level
3
AI Block personalized_tutor() generates adaptive content in local language
4
AI Block progress_assess() tracks mastery and identifies gaps
5
Progress data written to education registry with consent
6
Teacher notified of at-risk students via dashboard
7
Human teacher retains authority over promotion/assessment decisions
AI Blocks Used
personalized_tutor()progress_assess()language_adapt()
DPI Dependencies
Education RegistryStudent Digital IDData Exchange
Channels
Mobile AppOffline-capableTeacher Dashboard
Governance note: Student performance data is sensitive. Consent, purpose limitation, and data minimisation must be explicitly encoded in the DPI Workflow. AI informs — teachers decide.
Crisis Response — Emergency Benefit Disbursement
1
Disaster declared — affected geographic area defined by authorities
2
AI Block area_classify() identifies affected population from registries
3
AI Block eligibility_verify() rapid-matches against social protection registry
4
AI Block fraud_detect() screens for duplicate or anomalous claims
5
Human authority reviews and approves disbursement batch
6
Bulk payment triggered via existing payments infrastructure
7
SMS/WhatsApp notifications sent; audit log created for accountability
AI Blocks Used
area_classify()eligibility_verify()fraud_detect()
DPI Dependencies
Digital IDSocial RegistryPayments RailGIS Data
Channels
SMSWhatsAppUSSD
Key enabler: Pre-built, tested DPI Workflows for crisis disbursement can be activated rapidly. Countries without this infrastructure lose weeks standing up systems when it matters most.
Implementation Playbook

From architecture to deployment

9 sequential steps — each producing a concrete delivery output. Later steps rely on decisions and artifacts produced earlier. Skipping steps typically results in weak governance, fragile automation, or unclear accountability.

01

Identify the Use Case & User Journey

Start with one narrowly defined service journey that can be completed end-to-end. Map it in plain language before any technical work begins.

02

Identify AI Blocks, DPI Dependencies & DPGs

For each step in the journey, identify the minimum set of AI capabilities required. Each capability must be callable as a block, invoked only through the workflow.

03

Define Governance, Safeguards & Human-in-the-Loop

Before automation: owning ministry, legal mandate, consent requirements, escalation rules, audit and retention policies. Human oversight is not optional.

04

Create the DPI Workflow

Translate the user journey into a machine-readable DPI Workflow. Specify ordered steps, safeguards between steps, fallback logic, and authoritative outputs.

05

Define the Public Agent

Configure as a constrained interface. Supports local languages, can activate only approved workflows, cannot access data without consent, always offers human escalation.

06

Define Metrics & Operational Signals

Metrics must be defined before launch: success/failure rates per step, escalation frequency, language coverage, confidence scores, end-to-end completion time.

07

Implement the Service

Stand up the actual system. Code-based (APIs/SDKs) or No-Code/Low-Code (OpenFn, N8N). Both paths produce the same outcome. Start in supervised, human-in-the-loop mode.

08

Test & Validate

Synthetic end-to-end tests, AI Block accuracy validation against declared thresholds, safeguard validation, and human-in-the-loop dry run. No citizen contact before validation.

09

Fine-Tune & Improve Iteratively

Adjust confidence thresholds, expand language coverage, reduce manual escalation where safe, improve data mappings. All improvements via blocks and workflows — not ad-hoc logic.

Implementation Sandbox

See the framework in action

Watch a live end-to-end demo of a farmer requesting a government benefit via WhatsApp — with every AI Block, DPI system, and governance step visible in real time. Then assess your readiness, explore scenarios, and generate a roadmap.

Open the Sandbox →
🌾
Live Demo
Farmer benefit via WhatsApp — animated end-to-end
📊
Readiness Assessment
16 questions across 5 dimensions including AI adoption
🗺️
Roadmap Generator
Personalised 3-phase implementation roadmap
🔍
Scenario Explorer
5 sector implementation journeys step by step
📊

AI-DPI Readiness Assessment

Answer 14 questions across four dimensions to understand your country's readiness and get tailored recommendations. Takes about 5 minutes.

Explore an Implementation Scenario

Select a use case to walk through the step-by-step implementation journey, including the AI Blocks you'll need and the governance decisions required at each stage.

🛡️
Social Services
Benefits eligibility & disbursement
📝
Civil Registration
Birth certificates & identity
🌾
Livelihoods
Farmer benefits via WhatsApp
📚
Education
Personalised learning support
🚨
Crisis Response
Emergency benefit disbursement
🗺️

Generate Your Implementation Roadmap

Complete the Readiness Assessment first to get a personalised 3-phase implementation roadmap tailored to your country's gaps and strengths.

Or generate a generic roadmap:

Country Archetype Comparison

See how your readiness profile compares to four common country archetypes. Complete the assessment to overlay your profile, or explore the archetypes directly.

Resources

Read, cite, and implement

📄 Vision Paper

DPI-AI Framework

The full framework paper: executive summary, architecture, use cases, challenges, glossary, and references.

📋 Implementation Playbook

From Architecture to Deployment

Step-by-step playbook with YAML templates, AI Block specifications, and governance checklists for running a pilot.

📖 Full Paper (Alt)

The DPI-AI Framework

Extended version of the framework paper with detailed architecture diagrams, patterns, and implementation guidance.

🏛️ About CDPI

Centre for Digital Public Infrastructure

CDPI works to accelerate the adoption of DPI globally, providing technical assistance, convening, and knowledge resources.

Visit cdpi.dev →
📖 How to Cite

Citation

CDPI Abadie, D. (2026). DPI-AI Framework: Building AI-Ready Nations through Digital Public Infrastructure.

https://digitalpublicinfrastructure.ai →
⚖️ License

Creative Commons CC BY 4.0

This work is licensed under Creative Commons Attribution 4.0 International. Free to share and adapt with attribution.

View license →