Validating an AI-Driven Personalized Learning Framework for SDG 4: A Design Science Approach in LMIC Contexts

In alignment with Sustainable Development Goal 4 (SDG 4), this study proposes and validates a theoretical framework for an AI-supported personalized learning system designed for Low- and Middle-Income Countries (LMICs). Following a Design Science Research Methodology (DSRM), the framework was refined through a two-round Delphi study involving experts in AI, education, and policy. Reliability was further assessed through repeated simulations using the EdNet, OULAD, and Khan Academy datasets to model learner diversity and path adaptability. Results indicate that the refined model successfully maintains instructional stability across varied digital infrastructure scenarios while mitigating algorithmic bias, providing a scalable foundation for future educational R&D.