Presenting her cutting-edge research to multiple Global Summits in Silicon Valley and New York, UCL School of Management Associate Professor Anglea Aristidou is emphasising the social benefits of AI for Social Good (AI4SG) on the global stage.
Since the start of 2018, at least 35 major funding initiatives have directed more than $410 million towards AI for Social Good projects. As investment continues to accelerate, billions of dollars are expected to flow into AI for good initiatives, with much of that funding channelled through partnerships between frontier AI laboratories and nonprofit organisations.
Dr Aristidou and collaborators Vera Lummis and Rob Reich (Stanford) identified a significant gap in understanding how these AI for Social Good partnerships operate in practice, and why some generate lasting benefits while others fail to deliver sustainable impact.
In their recently published research, the team examined a series of real-world partnerships involving frontier AI labs, governments and nonprofit organisations working to deploy AI for social good. The study draws on interviews across 24 organisations in the AI4SG ecosystem, including philanthropic funders, and includes four in-depth case studies spanning health, education, agriculture, climate and humanitarian response.
The findings suggest that while AI partnerships have considerable potential to improve lives, success depends on far more than access to cutting-edge technology. The researchers identify the conditions that enable effective and sustainable collaborations, while also highlighting why many government and nonprofit organisations struggle to maintain AI solutions once subsidised support from technology companies comes to an end. The study also explores the role of open-source approaches and the influence of philanthropic foundations in shaping the wider AI4SG ecosystem.
Providing what the authors describe as the first empirical examination of frontier AI lab and nonprofit partnerships, the research offers evidence-based insights into a rapidly expanding field and has already attracted interest from AI practitioners, philanthropists and policymakers.
The findings were first presented on 15 September at the Imagination in Action Global Summit, a leading forum that brings together technology leaders, researchers and policymakers to discuss emerging innovations.
On 20 September, Professor Aristidou was invited to share insights from the research during events taking place alongside the 81st United Nations General Assembly, including the Global Philanthropic Leaders gathering at Goals House in New York City.
There, Professor Aristidou discussed the study’s findings with leaders from philanthropy, technology and civil society, contributing to conversations about how AI can be deployed more effectively, responsibly and sustainably in support of communities around the world.
Professor Angela Aristidou said: “There is enormous enthusiasm for using AI to tackle pressing social challenges, but our research shows that technology alone is never enough. The most successful partnerships are built on shared goals, strong governance and long-term investment in organisational capabilities.
“If we want AI to create lasting public value, we need to focus not only on innovation but also on building partnerships that enable communities and organisations to sustain that impact over time.”
By understanding what makes collaborations succeed or fail, the researchers hope their findings will help ensure that growing investment in AI leads to social benefits that are durable, inclusive and accountable.
Frontier AI Lab and Nonprofit Partnerships to Deploy AI for Good: Insights from an Empirical Study was published in September 2026.
