Bridging the Gap: Defining the Problem Space for AI-Driven Personalized Learning in Global Education

Achieving Sustainable Development Goal 4 (SDG 4) requires a fundamental shift in how educational resources are distributed and personalized, particularly in Low- and Middle-Income Countries (LMICs). This paper defines the problem space for AI-driven personalized learning by conducting a systematic review of existing technologies and their limitations in resource-constrained settings. Through a synthesis of literature (2013–2024) and an analysis of large-scale datasets including EdNet (131M interactions), OULAD, and Khan Academy, this research identifies a critical gap: the lack of scalable, context-aware AI frameworks that prioritize cultural localization and low-resource resilience. The findings suggest that while AI offers transformative potential for engagement, current solutions remain hindered by infrastructure dependency and algorithmic bias.