McGovern Foundation’s AI grants prioritize institutions over breakthroughs

The Patrick J. McGovern Foundation, a philanthropy focused on advancing artificial intelligence for public purpose, has unveiled more than $75 million in AI grantmaking this year aimed at shaping how the technology is actually used — and by whom. The foundation’s 2025 funding includes 149 grants across 13 countries. Recipients range from Digital Green, which is scaling AI-enabled agricultural advisory services for smallholder farmers, to Direct Relief, which is applying machine learning to improve forecasting in humanitarian medical response. The United Nations Office for Digital and Emerging Technologies also received support to develop blueprints for centers to help bridge the AI divide. The announcement pulls together the foundation’s AI grants made over the past year, offering a snapshot of where it is directing funding and why. These investments bring the McGovern Foundation's total grantmaking to $500 million. The focus is less on technological novelty than on ensuring communities and public institutions can work together to build AI solutions that meet real needs. In an interview with Devex, Vilas Dhar, president and trustee of the McGovern Foundation, said the next phase of AI’s impact will depend less on the pace of technical breakthroughs than on whether institutions are equipped to guide how those technologies are deployed. “In so many places, AI is already arriving, but it's arriving through mechanisms that aren't connected to community,” he said. “It's coming through procurement decisions, pilots that are funded by philanthropies, or global institutions that are coming in saying, ‘We're going to give you compute and data centers.’” Without the right institutional capacity for responsible AI, Dhar warned, those investments risk remaining fragmented and disconnected from the communities they are meant to serve. Dhar spoke with Devex on Monday from New York’s John F. Kennedy International Airport, shortly after returning from Trinidad and Tobago, where the foundation helped launch the Caribbean Artificial Intelligence Innovation Centre — one example of its work with governments, civil society organizations, and multilateral institutions on AI governance. Increasingly, AI companies are getting involved in grantmaking. Earlier this month, for example, OpenAI announced more than $40 million in grants for U.S.-based nonprofits. “Companies are racing for ever more powerful models that have certain fundamental capacities, but those capacities aren't translated into the layer of applications that actually serve public purposes,” Dhar said. He pointed to AI models capable of understanding how to intervene against the spread of antibiotic-resistant tuberculosis. Turning that technical capability into real-world impact, Dhar said, requires sustained investment in institutions, data systems, and delivery mechanisms needed to apply existing tools to public health challenges — areas where private capital is often poorly aligned with development needs. “We need a new set, a layer of institutions, that are able to take the incredible innovation happening in these companies and turn them into real outcomes,” Dhar said. Philanthropy, he added, can act as risk capital — backing speculative use cases early while also supporting the organizations needed to apply and distribute AI tools at scale. “Of course, we welcome future philanthropic investment from the large companies, but what they're able to bring to the table is technological knowledge and capacity, but very little in the way of learned and lived experience of how to solve complex, systemic problems,” he said. “So what we actually need is, of course, massive engagement from the private sector, but then we need philanthropy, civil society, and development organizations to bring forward clear encapsulations of problems and innovative approaches that leverage AI.” From JFK airport, Dhar boarded a 15-hour flight to New Delhi for meetings related to the India–AI Impact Summit that will take place there in February. “It’s shaping up to be maybe the most important conversation on how we actually align public and private capital to actually design for community needs,” he said. For Dhar, the risk is that AI funding remains overly focused on short-term returns, often prioritizing AI pilots and use cases. He hopes to see more funders take a medium- and long-term approach — involving society early, supporting institutions that ensure AI benefits humanity, and investing in the governance structures needed to guide the technology’s impact. “We might have in 10 years some of the most incredible and advanced AI systems ever built, but have them so far away from where people actually live and work and find economic dignity that they don't matter for billions of people,” he said.