Logo Lanfrica
  • Home
  • Atlas
  • Insights
  • Docs
  • Sign in

© 2026 Lanfrica. All rights reserved. All copyrights of the resources shown on the Lanfrica website belong to the original copyright holders, unless explicitly stated otherwise.

AI Policy Design and Implementation Toolkit: Twenty-five connected instruments for the ACTION policy lifecycle

Domain:

digital infrastructure

Record type:

toolspolicy
Creator:
NajHayHarIsa
Editor:
NgcIsaAliSul
Publisher:
Zenodo
Host:avatar
A working instrument set for public officials designing, appraising, procuring, piloting and overseeing artificial intelligence in the public sector. It is the practitioner companion to the National AI Governance Blueprint for the Republic of Ubuntia: where the blueprint sets out what a national AI governance framework should contain, this toolkit provides the forms through which officials reach and record the decisions that framework requires. Twenty-five instruments are grouped into six modules and mapped to the six stages of the ACTION policy lifecycle — Assess, Compare, Translate, Implement, Oversee, Navigate: Define and diagnose — problem definition, context and readiness, stakeholder mapping, evidence register Assess policy options — AI necessity and alternatives, options appraisal, risk and opportunity, distributional impact Design governance — rights and inclusion impact, data governance, institutional mandate, public participation Procure and prepare — responsible procurement, vendor and dependency, implementation readiness, capacity and resources Pilot and implement — pilot charter, roadmap, human oversight and remedy, public communication Monitor, adapt and exit — monitoring and evaluation, incident reporting, material-change review, the scale-revise-pause-stop scorecard, exit and transition Each instrument states the decision it supports, the completion standard it must meet, the gate rule that governs whether work may proceed, and the errors most often made in using it. Three assessment pathways — rapid, standard and enhanced — scale the effort to the consequence, with automatic escalation triggers. The toolkit is governed by a non-compensable rule: no aggregate score may override missing lawful authority, decision-critical evidence, meaningful affected-person participation, credible remedy or a practical exit route. Completed instruments assemble into a single AI Policy Action File. The instruments were developed from the eight capstone projects submitted by Cohort 1 of the AI Literacy Fellowship for African Policymakers, written by forty-one fellows and published in full in the companion volume. Worked examples use the Republic of Ubuntia, a composite country invented for the fellowship. Published by the OpenSchool Initiative, Abuja, Nigeria. First open publication edition, 2026. Licensed CC BY 4.0. This is an open educational and policy-learning resource. Completing these instruments confers no legal, procurement or clinical authority and does not authorise the deployment of an AI system.

Visit

doi.org

Licenses

Creative Commons Attribution 4.0 Internationalhttps://creativecommons.org/licenses/by/4.0/legalcode© 2026 OpenSchool Initiativehttp://rightsstatements.org/vocab/InC/1.0/