Organizations Working on Universal AI Access
The vision outlined in Chapter 14 — bringing AI to the 2.2 billion people without internet access — isn't starting from scratch. Organizations around the world are already working on pieces of this puzzle: digital infrastructure, offline technology, AI in local languages, community access centers, and more.
This page catalogs organizations you can support, partner with, or learn from.
Major International Initiatives
These organizations coordinate large-scale efforts across governments, corporations, and civil society.
EDISON Alliance (World Economic Forum)
What they do: A coalition of 200+ partners that has connected over 1 billion people to essential digital services in healthcare, education, and finance across 100+ countries — ahead of their 2025 target.
Why it matters: Proof that coordinated public-private partnerships can achieve scale. Their "Lighthouse Countries Network" (Ethiopia, Ghana, Nigeria, Rwanda, and others) provides models for national digital inclusion programs.
- Website: edisonalliance.org
- Impact Report: EDISON Alliance Impact Report 2024
International Telecommunication Union (ITU)
What they do: The UN agency for information and communication technologies. Sets global standards, coordinates spectrum allocation, and runs the World Telecommunication Development Conference (WTDC) that shapes development priorities for 4-year cycles.
Why it matters: The ITU's data on global connectivity (the 2.2 billion figure comes from them) and their Broadband Commission for Sustainable Development set targets that governments and organizations work toward.
- Website: itu.int
- Key Resource: ITU Facts and Figures 2025
Internet Society
What they do: Supports Internet infrastructure development, particularly in least-developed countries, small island developing states, and landlocked developing countries. Focuses on community networks, Internet Exchange Points, and policy advocacy.
Why it matters: They work on the foundational layer — getting connectivity infrastructure built where it doesn't exist.
- Website: internetsociety.org
World Bank Digital Development
What they do: Finances digital infrastructure projects in developing nations. Their Knowledge for Change Program specifically funds research on "AI for Development" as a priority area.
Why it matters: The World Bank is often the largest single funder of infrastructure in developing countries. Their priorities shape what gets built.
- Website: worldbank.org/en/topic/digitaldevelopment
- Funding: Knowledge for Change Program
Digital Literacy and Education
Organizations bringing reading, learning, and digital skills to underserved populations.
Worldreader
What they do: A global nonprofit that has supported 22+ million readers in 100+ countries since 2010. Their BookSmart app offers 3,000+ books in multiple languages (English, Swahili, Arabic, Hindi, Spanish) for children ages 3-12.
Why it matters: They've solved the "offline-first" problem — their app works without constant connectivity. Their partnership model (working with Ministries of Education, local NGOs, corporate partners like Safaricom) is replicable.
Impact: Studies show reading speed and comprehension significantly increased after just five months of use. Won the 2023 Library of Congress Literacy Award.
- Website: worldreader.org
- Where they work: Ghana, Kenya, Uganda, South Africa, Egypt, India, and more
Khan Academy
What they do: Free, world-class education for anyone, anywhere. Increasingly used in Global South countries and has been developing offline capabilities.
Why it matters: Demonstrates that high-quality educational content can be delivered at zero marginal cost. Their content is being used by many other digital inclusion initiatives.
- Website: khanacademy.org
African AI and Language Technology
Organizations building AI that actually works for African languages and contexts.
Masakhane
What they do: A grassroots organization building Natural Language Processing (NLP) tools for African languages. "Masakhane" means "We build together" in isiZulu. Over 1,000 contributors from 30+ African countries have created machine translation benchmarks for 52+ African languages.
Why it matters: This is exactly what Chapter 14's vision requires — AI that speaks local languages. Masakhane proves it can be done by Africans, for Africans, with community ownership of data.
Key principle: Data sovereignty — Africans should decide what data represents their communities and retain ownership of it.
- Website: masakhane.io
- GitHub: github.com/masakhane-io
- How to help: Train models, contribute datasets, provide compute resources, or mentor
Ghana NLP
What they do: Developing NLP tools specifically for Ghanaian languages, working closely with the Masakhane community.
Why it matters: Country-specific focus allows deeper engagement with local linguistic and cultural context.
- Website: ghananlp.org
Lanfrica
What they do: A directory and repository of African NLP datasets, making it easier for researchers and developers to find and use African language data.
Why it matters: Addresses the critical data gap — you can't build AI for a language without data in that language.
- Website: lanfrica.com
Agricultural AI
AI helping the farmers who feed the world.
Farmerline (Ghana)
What they do: Dubbed the "Amazon for African farmers," Farmerline provides smallholder farmers with access to fertilizer, seeds, climate-smart farming education, and market connections. Their Darli AI chatbot provides instant farming advice in 27 languages (including Twi, Swahili, Yoruba) via WhatsApp.
Why it matters: This is the closest existing model to Chapter 14's vision:
- Works in local languages
- Accessible via WhatsApp (low bandwidth)
- Works without smartphones — farmers can call or text an AI helpline
- Already reaching 2.3+ million farmers across 50 countries
Impact: 30% productivity boost for farmers using the platform. Supported by the Green Climate Fund.
- Website: farmerline.co
- Countries: Ghana, Cote d'Ivoire, Togo, Benin, Burkina Faso, expanding across West Africa
Opportunity International — FarmerAI
What they do: AI-powered tools for smallholder farmers, focusing on financial inclusion alongside agricultural support.
Why it matters: Combines agricultural AI with microfinance, addressing multiple barriers simultaneously.
- Website: opportunity.org
Healthcare AI
AI improving health outcomes in underserved communities.
PATH
What they do: A global health nonprofit headquartered in Seattle, working to advance health equity. They've appointed a Chief AI Officer and are integrating AI to improve diagnostic accuracy, streamline health supply chains, and guide disease surveillance.
Why it matters: PATH has deep relationships with health ministries in developing countries. When they adopt AI tools, they bring implementation expertise and trust relationships.
- Website: path.org
Operation Smile
What they do: Provides free cleft lip and palate surgeries worldwide. In collaboration with Microsoft, they've integrated AI to analyze surgical photographs and train surgeons in remote areas.
Why it matters: Demonstrates AI as a tool for training and quality assurance in healthcare, not just diagnosis.
- Website: operationsmile.org
Connectivity Infrastructure
The physical layer — getting internet to places that don't have it.
Satellite Internet
| Project | Organization | Status |
|---|---|---|
| Starlink | SpaceX | Operational in 70+ countries, expanding to remote areas |
| Project Kuiper | Amazon | Launching 2025-2026, aims for global coverage |
| OneWeb | Eutelsat OneWeb | Operational, focus on enterprise and government |
Why it matters: Low Earth Orbit (LEO) satellite constellations can provide internet to areas where laying fiber or building cell towers is impractical. Cost is dropping rapidly.
Undersea Cables
| Cable | Organization | Route |
|---|---|---|
| 2Africa | Meta | Circles entire African continent, 46 landing points |
| Equiano | Europe to South Africa via West Africa |
Why it matters: These cables dramatically increase bandwidth capacity to the African continent, reducing costs and enabling services that weren't previously feasible.
Community Technology Centers
Shared spaces providing community access to digital technology.
The Telecenter Model
What it is: Community spaces with computers, internet access, and trained facilitators. Originated in Scandinavia in the 1980s, spread globally in the 1990s-2000s as "telecenters," "community information centers," or "digital villages."
Current status: Research shows telecenters complement (not compete with) smartphones — complex tasks still benefit from larger screens, keyboards, and in-person support. Ghana's Community Information Centers significantly improved youth digital literacy and employment outcomes.
Key insight: The community center model that Chapter 14 proposes for AI access builds on decades of experience with telecenters. We know what works and what doesn't.
Challenges: Sustainability, reaching the poorest populations (not just emerging middle class), keeping technology current.
Resources
Research and Advocacy
Organizations studying digital inclusion and advocating for policy change.
Brookings Institution — Center for Technology Innovation
What they do: Policy research on AI governance, digital inclusion, and technology in the Global South.
- Key resource: AI in the Global South: Opportunities and Challenges
Access Now
What they do: Defends digital rights of users at risk around the world, including internet shutdowns, surveillance, and digital exclusion.
- Website: accessnow.org
The Partnering Initiative
What they do: Expertise in multi-stakeholder partnerships, including public-private-philanthropy partnerships (PPPPs) for development.
- Website: thepartneringinitiative.org
- Key resource: Introduction to Multi-Stakeholder Partnerships (PDF)
How to Get Involved
As an Individual
- Donate to organizations doing this work (Worldreader, Masakhane, and Farmerline all accept donations)
- Volunteer technical skills — Masakhane actively seeks contributors for model training, dataset creation, and mentorship
- Advocate for digital inclusion policies with your representatives (see our advocacy page)
As an Organization
- Partner with existing initiatives rather than starting from scratch
- Fund local organizations who understand their communities
- Share data responsibly to help build AI for underrepresented languages
As a Researcher or Developer
- Contribute to Masakhane — they need people to train models, annotate data, and build tools
- Design for offline-first — assume intermittent connectivity
- Prioritize local languages — English-only AI excludes billions
The Big Picture
The infrastructure for universal AI access is being built right now, in pieces, by organizations around the world. What's missing is coordination and scale.
Chapter 14 of My Adventures With Claude lays out a vision for how these pieces could come together — community AI centers, offline-capable models, voice interfaces in local languages, multi-stakeholder funding.
The organizations on this page prove it's not a fantasy. The technology exists. The models work. The question is whether we choose to deploy them at the scale the world needs.
Return to AI for All advocacy page or Chapter 14: Universal AI Access.