Leveraging artificial intelligence and big data analytics for poverty alleviation and social development: a review

Abstract

Poverty reduction remains one of the most pressing global development challenges, requiring innovative approaches for identifying vulnerable populations, targeting interventions, and monitoring outcomes. Rapid advances in Artificial Intelligence (AI), Machine Learning (ML), and Big Data Analytics (BDA) have created new opportunities to complement conventional poverty assessment methods, which often rely on costly, infrequent household surveys that are susceptible to measurement errors and reporting biases. AI-driven approaches leverage high-resolution satellite imagery, mobile phone metadata, digital financial transactions, administrative records, and social media data to generate timely and spatially detailed insights into poverty dynamics. This study presents a systematic literature review of publications from 2014 to 2024, supplemented by recent policy reports and selected case studies, to examine the application of AI and BDA in poverty alleviation across finance, agriculture, healthcare, education, and social protection. The review further evaluates ethical, governance, and institutional issues through the World Bank's 4Cs AI readiness framework, comprising connectivity, compute, context, and competency. Findings indicate that AI and BDA have significant potential to enhance poverty measurement, improve service delivery, and support evidence-based policymaking. However, their long-term effectiveness depends on responsible deployment, inclusive governance, robust data protection, investment in digital infrastructure and human capacity, and locally adaptive implementation strategies. Unlike previous sector-specific reviews, this study provides a comprehensive cross-sectoral synthesis integrating technological, ethical, governance, and developmental perspectives for Low- and Middle-Income Countries (LMICs).