This study aims to identify patterns of application, barriers to implementation, and recommendations for applying artificial intelligence (AI) in education across developed and developing countries, considering AI’s dual potential to bridge or widen the global education gap. Using a Systematic Literature Review (SLR) guided by PRISMA 2020 guidelines, articles were systematically searched in Scopus, ERIC, and ScienceDirect databases for the period 2015-2025. After applying predefined inclusion and exclusion criteria, 63 empirical studies conducted in 45 countries were selected for analysis. The findings show a significant increase in global research interest in AI in education, particularly after 2024. AI implementation patterns in education are primarily categorized into three functions: AI as a teaching and learning tool, AI as a learning analytics instrument, and AI as an adaptive learning system. The level of implementation differs between country contexts; developed countries generally demonstrate integrated and transformative adoption, while developing countries remain largely at exploratory and supportive stages. Major barriers reported in the literature include limitations in infrastructure, funding, and human resource capacity, as well as ethical concerns, governance challenges, and insufficient policy frameworks. These findings highlight that the impact of AI on educational equity is context dependent. When supported by inclusive policies, adequate infrastructure, and improved AI literacy among educators and students, AI can enhance access to learning and reduce disparities. However, without equitable access and strategic governance, AI adoption risks reinforcing existing educational inequalities between and within countries.
Budiman et al. (Tue,) studied this question.