Skip to main content
Lampas HumanityGlobal Public Technology Initiative
Lampas HumanityGlobal Public Technology Initiative

Technology should empower humanity. Progress should benefit everyone.

Initiative 02 · Proposed

Universal AI Infrastructure

Affordable AI access, open-source intelligence, shared computing infrastructure, education and public research.

Proposed initiativeBridges Technology and Communities
02

Introduction

Access to artificial intelligence is becoming a determinant of economic opportunity, scientific progress and the quality of public services. Universal AI Infrastructure is a proposed effort to make capable, trustworthy AI affordable and accessible to public institutions, researchers, educators, small enterprises and individuals — particularly where access is limited today.

Objectives

What it would aim to achieve.

  1. 01Lower the cost of accessing AI capabilities for public-interest uses.
  2. 02Support open-source models, tools and datasets that can be inspected, adapted and improved.
  3. 03Develop shared computing arrangements for researchers, educators and public institutions.
  4. 04Expand AI education and skills for students, teachers, public servants and workers.
  5. 05Encourage AI development in local languages and for local needs, with the consent of the communities involved.

Problems addressed

Why it is needed.

  • Computing power and advanced AI capabilities are concentrated among a small number of organisations and countries.
  • Many languages and contexts are underrepresented in AI systems, which limits their usefulness and fairness.
  • Public institutions and researchers in many regions cannot afford the infrastructure needed to build or evaluate AI.
  • AI literacy is uneven, leaving many people unable to use AI effectively or to scrutinise how it affects them.

Proposed technological solutions

How technology could help.

These are candidate approaches to be tested in research and pilots — not finished tools. Each would be subject to the safeguards set out in our governance framework.
  • 01

    Shared computing access

    Cost-recovered access to computing resources for public-interest research and education, allocated through transparent criteria.

  • 02

    Open model and tool catalogue

    A curated catalogue of open-source models and tools, with documented capabilities, limitations and known risks.

  • 03

    Language and context inclusion

    Support for datasets and models in underrepresented languages, built with community consent and fair sharing of benefits.

  • 04

    Open curricula and training

    Openly licensed learning programmes in AI literacy and applied AI skills, designed with educators.

  • 05

    Independent evaluation

    Methods for testing AI systems for safety, accuracy and bias in the contexts where they will actually be used.

  • Safeguards apply

    Every approach is subject to impact assessment, independent review and public reporting.

    The governance framework

Expected public benefits

Who stands to gain.

The intended benefits, if the initiative is developed and proves effective. They will be measured through independent evaluation.
  • Broader participation in AI research and development.
  • AI tools that better reflect the languages and needs of the people using them.
  • Lower barriers for public institutions seeking to use AI responsibly.
  • A workforce better prepared to work with — and to scrutinise — AI systems.

Future development roadmap

A phased path, without fixed dates.

Each phase would begin only when the previous one justifies it. Timelines will be set — and published — once the necessary partnerships and governance are in place.
  1. Phase 1

    Research and design

    Map existing open-source models, computing resources and training programmes. Identify the gaps that matter most for public-interest users.

  2. Phase 2

    Partnership and pilot design

    Explore arrangements with universities, research institutions and technology providers for shared computing access and training.

  3. Phase 3

    Initial pilots in Africa

    Pilot shared access and training programmes with interested African universities and research groups.

  4. Phase 4

    Scaled, open deployment

    Expand shared infrastructure and open resources internationally, governed by transparent access and acceptable-use policies.

Within the ecosystem

Provides analytical capacity to Global Development Intelligence and skills pathways to the Global Opportunity Exchange, under the safeguards set by Digital Rights and Protection.

  • 05Proposed

    Digital Rights and Protection

    Privacy, cybersecurity, digital identity, fraud detection, whistleblower protection and responsible data governance.

  • 01Proposed

    Global Accountability Network

    Technology for public financial transparency, procurement monitoring, institutional accountability, public project tracking and citizen reporting.

  • 03Proposed

    Global Opportunity Exchange

    Employment, skills development, entrepreneurship, manufacturing, business connections and economic opportunities.

  • 04Proposed

    Community Ownership Network

    Technology supporting transparent community participation in productive assets, infrastructure and legally compliant investment.

Contribute

Help shape the Universal AI Infrastructure.

We welcome expertise, critique and collaboration at every stage — especially the earliest.