In 2025, African start-ups raised roughly $4 billion in venture funding across the entire continent. That same year, one San Francisco–based artificial intelligence (AI) company raised ten times that amount.
Health technology claims a thin slice of the continental figure, and AI for health’s piece is thinner still. Meanwhile, Africa carries about a quarter of the world’s disease burden with only 3 percent of its health workforce. The region is in dire need of innovative solutions to address its health worker shortage, but that financing gap is a structural constraint on who can build the tools that will define the next era of health care.
Philanthropy ends up filling some of the gap, which is not inherently negative, given that catalytic philanthropic capital has seeded some of the world’s most inventive health and AI companies. But philanthropic capital behaves nothing like a market. Grants end. When they do, the domestic financing that should take over in low- and middle-income countries is rarely ready: Government budgets, insurance schemes, and reimbursement mechanisms are seldom positioned to finance, even when the company or technology is delivering significant public health value.
Too many African innovators succeed at the pilot and stall at the procurement, having scaled on a class of capital that was never designed to sustain them.
It has been a privilege, both prior to and during my time at PATH, to have advocated, loudly and often, for more risk capital to flow to African innovators. Advocacy alone can treat the symptoms—it can encourage more funds to flow in—but it’s not getting at the root problem: what makes these markets so difficult to invest in? The answer has little to do with the quality of the companies themselves. It is the absence of enabling infrastructure that prevents a good product from becoming a scalable business.
The US health system offers an instructive example. The American Medical Association maintains the CPT (Current Procedural Terminology) coding system, a comprehensive catalogue of clinical activity that lists every diagnosis, procedure, and intervention, including services delivered through medical devices and, increasingly, AI-enabled tools. Once a technology has a code, any payer, whether a public program or a private insurer, can attach a price to it. A developer no longer negotiates site by site, with separate contracts and separate usage monitoring at every hospital. One administrative artifact collapses thousands of sales conversations into a handful of negotiations. It is profoundly boring infrastructure, and yet it is one of the principal reasons that AI technology companies have found traction in the United States.
The global equivalent for clinical terminology is similarly unglamorous. SNOMED CT, the world’s largest collectively owned clinical ontology, is comprehensive enough to record a waterskiing mishap, an accident at an atomic power plant, and, should medicine ever require it, an accident involving a spacecraft. Yet only a handful of African countries have implemented SNOMED CT nationally (South Africa officially joined just last month).
The rails that let a Boston diagnostics company sell into 50 states do not run between Kigali, Lagos, and Accra. Thus, most African developers with a clinically validated AI tool face a different journey: bespoke integration with every facility’s records, bespoke contracting with every purchaser, bespoke …, bespoke …, bespoke. The cost of that friction lands precisely on the companies least able to absorb it.
This infrastructure matters in both directions. An innovator in Kampala who builds against shared terminologies and coding standards from day one is building for continental and global markets, ready to plug into reimbursement systems wherever they exist. Governments that adopt those standards gain something equally valuable: the ability to import best-in-class technologies when importing is the right answer, and to verify, monitor, and pay for what they buy. Interoperability, done well, is sovereignty’s quiet ally.
This is how PATH approaches problems under our new 2030 strategy, which places as much weight on the how of impact as on the what.
We start with the people closest to the problem. Our regional AI, data, and digital health innovation centers—currently in Africa and Asia-Pacific—exist for exactly these purposes: to sit with innovators, regulators, and ministries of health; to understand what actually inhibits them; and to let those constraints shape the work we do together. Some of what emerges will be advocacy for capital. Much of it will be the assembly of coding systems, tariff structures, and procurement pathways for which no ribbon will ever be cut.
The next decade of African health innovation will be decided as much by terminologies and tariff schedules as by investors and engineers. Markets capable of sustaining local innovation are built deliberately, by partners willing to do the unglamorous work of laying rails.
PATH intends to be that partner, and we are already at work.
