Evidence for AI in Health (EVAH) Initiative Research Grants is sponsored by Gates Foundation, Novo Nordisk Foundation, and Wellcome. The Evidence for AI in Health (EVAH) Initiative is a joint US$60 million program from the Gates Foundation, Novo Nordisk Foundation, and Wellcome to fund locally led evaluations of AI-enabled health tools in low- and middle-income countries. Application snapshot: target deadline April 1, 2026; published funding information See official notice; eligibility guidance Open to researchers globally, including those not affiliated with J-PAL. Use the official notice and source links for final requirements, attachment checklists, allowable costs, and submission instructions before applying.

Evidence for AI in Health (EVAH) is sponsored by Bill and Melinda Gates Foundation (joint with Novo Nordisk Foundation and Wellcome Trust). Open call for proposals evaluating mature AI tools in primary and community health settings in low- and middle-income countries. Application snapshot: target deadline April 1, 2026; published funding information Up to $1,000,000 (Pathway A) or up to $3,000,000 (Pathway B); eligibility guidance Organizations with access to an AI-enabled clinical decision support tool deployed or ready for deployment in primary health care in Sub-Saharan Africa, South Asia, or Southeast Asia; nonprofits, for-profits, internatio… Use the official notice and source links for final requirements, attachment checklists, allowable costs, and submission instructions before applying.

The Evidence for AI in Health (EVAH) Initiative is a joint US$60 million program from the Gates Foundation, Novo Nordisk Foundation, and Wellcome to fund locally led evaluations of AI-enabled health tools in low- and middle-income countries. The initiative supports rigorous research including randomized controlled trials, implementation science studies, economic analyses, and assessments of public acceptance of AI health tools. Studies focus on AI-enabled decision support tools for frontline health workers performing clinical tasks such as triage, diagnosis, and referral in primary and community health settings. All findings are shared through open-access channels prioritizing data privacy and ethical principles. The initiative addresses a critical evidence gap: only four of 86 identified AI health trials between 2018 and 2023 occurred in LMICs. Applications are due April 1, 2026 through the J-PAL hosted portal.

The Wellcome Trust funds research applying generative AI to mental health, specifically anxiety disorders, depressive disorders, and psychotic disorders. Awards of up to £3 million (approximately USD $3.8 million) support up to 2 years of research following a 4-month accelerator stage. The program supports two research areas: (1) creating or improving generative AI models for mental health measurement and intervention, and (2) generating evidence on safe collaboration between generative models, mental health professionals, and people with lived experience. The program explicitly does not fund real-world deployment or broader population applications, focusing instead on fundamental research. Teams receive accelerator support before the full grant phase.

Open Philanthropy's landmark Technical AI Safety Research RFP is one of the largest dedicated funding programs for AI safety research, committing approximately $40 million across 21 research directions organized into five clusters. The clusters include: (1) Adversarial Machine Learning — jailbreak detection, control evaluations, and alignment stress tests; (2) Sophisticated Misbehavior — alignment faking, deceptive behavior, and hidden capability detection; (3) Model Transparency — interpretability applications, activation monitoring, and feature discovery; (4) Trust from First Principles — rare misbehavior estimation and theoretical inductive biases; (5) Alternative Approaches — conceptual clarity on AGI risks and novel alignment agendas. The application process begins with a streamlined 300-word Expression of Interest (EOI), with responses within two weeks. Promising EOIs are invited to submit full proposals, with funding decisions approximately two months later. The program explicitly supports research on large language models or work with transferable insights to LLMs. Administered through Coefficient Giving, the fund draws on contributions from Good Ventures and over 20 additional philanthropists.

The Coefficient Giving AI Governance RFP funds research and projects addressing catastrophic risks from advanced AI through governance mechanisms. The program supports six primary subject areas: (1) Technical AI governance including compute governance, model evaluations, and standards development; (2) Policy development for AI regulation and oversight; (3) Frontier company policies to improve internal safety and governance practices; (4) International AI governance frameworks and cooperation; (5) Legal frameworks for AI accountability and liability; and (6) Strategic analysis and threat modeling of AI risks. Funded activities include research projects, training and mentorship programs, general organizational support, and other innovative projects. The program is open to applicants worldwide from academia, nonprofits, industry, and independent researchers. While the January 2026 deadline has passed, Coefficient Giving has indicated that future AI governance RFPs are likely as they continue proactive grantmaking in this area. The fund draws on contributions from Good Ventures and over 20 philanthropists.