Decolonizing AI in Africa: Scoping Review Data Extraction Matrix (PRISMA-ScR)

As Artificial Intelligence (AI) accelerates across Africa’s digital landscape, its epistemic foundations remain largely Eurocentric, often reproducing colonial hierarchies in data, design, and governance. Scholars increasingly call for a decolonized AI that reflects African values and contexts. This research responds to that call by exploring how African epistemologies (such as Ubuntu and communal intelligence) can shape the development of ethical, inclusive, and contextually grounded AI systems. This scoping review follows the PRISMA-ScR protocol, synthesizing literature from 2015–2025 across peer-reviewed, grey, and policy sources. Thematic analysis categorizes insights into three domains: (i) epistemic decolonization, (ii) ethical and governance frameworks, and (iii) indigenous-centered technological innovation.