Generative Artificial Intelligence Use and Adoption in African Higher Education among Students and Lecturers: A Systematic Review and Meta-Analysis

Background: Artificial intelligence (AI) tools are becoming increasingly common in higher education, but their use in African universities has not been quantitatively summarized. This study estimated the prevalence of AI use among students and academic staff in African higher education institutions. Methods: A systematic review and meta-analysis of quantitative studies was conducted in line with PRISMA 2020 guidelines. AI use was operationalised primarily as engagement with generative AI tools (e.g., ChatGPT), although some studies assessed broader AI tool use in academic contexts. Seven cross-sectional studies from five African countries, involving 1877 participants, were included. Random-effects meta-analysis with logit transformation and restricted maximum likelihood estimation was used, alongside subgroup and sensitivity analyses. Results: The pooled prevalence of self-reported generative AI (including ChatGPT and related tools) use was 79.7% (95% CI: 56.6% - 92.2%), with substantial heterogeneity (I2 = 97.4%). In subgroup analysis, the pooled prevalence was 87.1% (95% CI: 43.1% - 98.4%) among students and 64.2% (95% CI: 3.2% - 99.0%) among lecturers, although no statistically significant difference was observed (p = 0.2509). Sensitivity analyses showed that the overall findings were broadly stable. Conclusion: AI use appears to be widespread in the sampled African higher education settings, although estimates vary considerably across contexts. The pooled estimate should be interpreted as reflecting predominantly generative AI use, given the emphasis of most included studies on tools such as ChatGPT. These findings highlight the need for context-specific governance, capacity-building, and more consistent research to support responsible AI integration in African universities.