The Evidence for AI in Health (EVAH) initiative has been launched to generate robust, locally led evidence on AI tools in healthcare, particularly in low- and middle-income countries. Developed in partnership with the Gates Foundation and the Novo Nordisk Foundation, the initiative will invest US$60 million over three years to evaluate AI technologies and guide decisions on which innovations are ready for adoption and scale. EVAH is part of a broader US$300 million partnership supporting global health research and development.
AI tools hold promise for improving patient outcomes, but evidence about their effectiveness in real-world settings remains limited, especially for solutions designed and deployed in low-resource contexts. EVAH aims to strengthen local evaluation capacity and provide the data needed to ensure AI tools are safe, effective, equitable, and adapted to the communities they serve. Local validation is seen as critical to preventing AI from worsening health disparities.
The initiative will focus on tools that align with national health priorities and can be integrated into primary and community healthcare settings. Evaluations will cover AI applications such as prediction models for disease risk, computer vision for analyzing medical images, large language models to support clinical decision-making, and multimodal AI that combines different data types for patient risk assessment. Priority will be given to technologies designed for resource-limited settings and trained on representative local data.
EVAH will employ a range of evaluation methods, including implementation research, randomized controlled trials, economic evaluations, and acceptability studies to assess performance, cost-effectiveness, and user perception. The first calls for proposals will target AI-enabled decision support tools for frontline health workers in Sub-Saharan Africa and South and South-East Asia, assisting with triage, diagnosis, or referral.
The initiative is run in partnership with the Abdul Latif Jameel Poverty Action Lab and the African Population Health Research Centre, leveraging their experience in research, policy-relevant evidence generation, and local engagement. Findings will be published openly to inform governments, researchers, industry, and communities, promoting transparency, trust, and ethical AI deployment. EVAH emphasizes participatory research, ensuring that AI development is grounded in the needs of local populations, ultimately accelerating the safe and equitable use of AI in health systems.
