The Clinton Health Access Initiative, the Agency Fund, and Endless Network have launched the AI Commons for Global Health, a new, shared initiative to ensure that as AI transforms healthcare, it works for the patients who need it most.
The initiative emerged from a convening at the Rockefeller Foundation’s Bellagio Center in Italy, which brought together innovators from low- and middle-income countries, global technology companies, frontier AI labs, academia, global health implementing partners, multilateral institutions, and funders—all aligned around a shared concern: what will it take to ensure AI improves health outcomes around the world. This is a concern we believe extends well beyond the people in that room, and this initiative is designed to be shaped by a much broader community.
AI is changing what is possible in healthcare faster than almost anyone predicted. For health systems in low- and middle-income countries that have long been constrained by too few doctors and too few resources, that matters enormously. But only if the tools actually work in the settings where they are needed.
Because most AI tools are built on data from high-income contexts, they speak the languages, reflect the diseases, and assume the infrastructure of those places. When a developer in Ghana or Mozambique tries to develop a tool for their context, they have little to build on: sparse relevant data, no benchmarks designed for their patient population, few standards to follow. They often have to build from scratch, alone.
This is not an accident. It is what happens when innovation follows the path of least resistance—toward the places where the necessary inputs, like money and data, are already abundant. This means that without intentional intervention, wealthy health systems are likely to get better and better tools, while everyone else waits or makes do with tools that were never designed for them.
Global health has watched this happen before. Each wave of new technology, from internet-based health tools to mobile health apps, brings a wave of experimentation. Without effective coordination and proactive planning for scale, these efforts struggle to achieve lasting health impact. We cannot afford to repeat that arc with AI.
That is why this group of innovators, global health practitioners, and AI experts gathered in Italy: to figure out what it would take to do things differently.
The answer the group kept coming back to was that AI for global health needs shared, foundational inputs that make it possible for developers and procurers in low-resource contexts to create and use solutions that actually work in their countries.
What the AI Commons for Global Health is building
That is what the AI Commons for Global Health is working toward. Four working groups are now active, each designed to produce open, collaborative resources others can build on. The groups are focused on:
Data exchange: Creating a marketplace, grounded in fair value exchange, where researchers and developers can access data from low- and middle-income countries, and equipping those sharing and using data with tools to ensure data privacy and safety.
Clinical decision support architecture: A free design blueprint for AI-enabled clinical decision support tools, so innovators in low-resource contexts around the world can build on existing work rather than starting from scratch.
Evaluations: Practical tools to make it easier to rigorously test whether an AI tool actually works in a given context, before and after it is deployed.
Procurement: Reusable guidance to help governments identify and acquire AI solutions that genuinely fit their health systems.
Get involved
We are at the beginning. The AI Commons for Global Health will only be as strong as the community behind it, and we are actively looking for people who want to help shape this work. If you are building AI tools for low-resource health settings, working on evaluations, focused on data access, sharing, and governance, advising governments, or funding innovation in this space, we want to hear from you.
The tools that could transform health are being built right now. The question is whether they will be built for everyone.
To get involved, reach out to Amarynth Sichel, Associate Director – AI, at asichel@clintonhealthaccess.org
