This study systematically examines the use of artificial intelligence (AI) tools in higher education and compares the gains and challenges experienced by students and educators across the Global North and Global South. Using the PRISMA protocol, a systematic literature review was conducted on peer-reviewed empirical studies published between 2021 and 2025, retrieved primarily from the Scopus database with supplementary searches via Google Scholar. 71 studies met the inclusion criteria and were analysed using a structured data-extraction framework and inductive thematic analysis. Findings show that generative AI tools particularly ChatGPT dominate AI adoption globally, accounting for the majority of reported use across both regions. However, similar tools are employed in fundamentally different ways. Institutions in the Global North demonstrate deeper pedagogical integration, experimentation, and curriculum-level innovation, supported by stronger infrastructure and institutional policies. In contrast, Global South institutions rely largely on AI for foundational academic support such as writing assistance, summarisation, and language correction, reflecting structural constraints including limited infrastructure, affordability challenges, and weak governance frameworks. Across regions, reported benefits include improved efficiency, academic performance, and learner engagement, while persistent challenges centre on academic integrity, ethical concerns, overreliance, data privacy, and uncertainty due to inadequate institutional guidance. The study underscores that global access to AI tools does not translate into equitable educational outcomes. It contributes a comparative, equity-focused synthesis and advances AI adoption research by highlighting the need for context-sensitive policies, institutional capacity building, and inclusive AI literacy initiatives to prevent the reinforcement of existing global inequalities in higher education.
Olohunfunmi et al. (Thu,) studied this question.