����ɫ��

Artificial Intelligence is increasingly being adopted in efforts to address poverty through innovations such as algorithmic credit scoring, automated social service delivery, and predictive tools for resource allocation. These technologies are often celebrated for their potential to improve efficiency, expand access, and optimise decision-making at scale.

However, to ensure these tools genuinely support the people they are intended to help, it is essential to consider not only what AI can do, but how and for whom it is being implemented. Who defines the problems AI seeks to solve? Whose perspectives are included in the design of these systems? And how do we ensure that technological solutions remain grounded in the realities of the communities they aim to serve?

This talk explores AI not merely as a set of tools, but as part of a broader system shaped by social, institutional, and ethical considerations. Drawing on practical examples, I will examine how AI is reshaping definitions of “need” and influencing the distribution of support and resources. While these technologies hold promise, they also carry risks—particularly when they reduce complex human experiences to data points or overlook the nuances of daily life in underserved communities.

I will also highlight the critical importance of qualitative methods—such as interviews, ethnographic research, and community engagement—in complementing data-driven approaches. These methods help illuminate what is often invisible in datasets: context, meaning, and human experience.

As a solution, I propose a co-design model for AI in social systems—one that brings together technologists, policymakers, and the communities most affected by poverty. By creating inclusive design processes, supporting interdisciplinary research, and embedding accountability mechanisms into AI development, we can build systems that are not only effective, but equitable. The goal is not to reject the use of AI in poverty alleviation, but to advocate for more thoughtful, inclusive, and accountable systems—ones that are informed by the people most affected and grounded in a deep understanding of social realities.

Access the Zoom link here: