Generative AI to bridge the educational divide: Personalized learning and challenges
Aziz Mimoudi
Abstract
Generative artificial intelligence (GenAI) is rapidly transforming education by enabling adaptive tutoring, personalized feedback, and scalable resource creation. Yet its benefits are unevenly distributed: without safeguards, the same systems that promise to close learning gaps may reinforce them. This article presents a systematic review of 75 peer-reviewed studies published between 2016 and 2024, focusing specifically on GenAI through an equity lens. Four main opportunities were identified: personalized learning (60 % of studies), teacher assistance and efficiency (51 %), expanded access and inclusion (44 %), and learning analytics (33 %). Alongside these benefits, significant challenges were reported, including data privacy risks (56 %), algorithmic bias (52 %), depersonalization of learning (48 %), the digital divide (48 %), and transparency and accountability gaps (36 %). The synthesis reveals that GenAI's educational impact is shaped by three intersecting dimensions – personalization, fairness, and access – whose intersections determine whether GenAI narrows or widens learning inequalities. To achieve equitable outcomes, schools and ministries must prioritize governance measures such as privacy safeguards, bias audits, professional development, and infrastructure investment, since equity will not emerge automatically from GenAI adoption; it must be intentionally designed, implemented, and monitored across policy, practice, and pedagogy. Future research should move beyond descriptive accounts to include pragmatic trials in low-resource contexts, longitudinal equity studies, and governance frameworks that integrate technical safeguards with participatory oversight. By addressing these priorities, GenAI can shift from a tool of potential exclusion to a driver of equity and inclusion in education worldwide.