Decolonising AI: a Pan-African collaboration

Abstract

Amid the growing global adoption of Artificial Intelligence (AI) in higher education, concerns regarding its ethical use, cultural relevance, and epistemic consequences are becoming increasingly urgent, particularly within African contexts. This paper arises from a Pan-African academic collaboration that critically examines AI from a decolonial perspective, focusing on multilingualism, assessment, and pragmatic application in universities. The authors draw on decolonial theory and Afrocentric philosophy to promote the inclusion of African languages, epistemologies, and pedagogical values, while acknowledging how current AI systems reflect linguistic hierarchies embedded in global data infrastructures. Framed by an exploration of how cross-continental collaboration can advance contextually grounded, socially responsive, and inclusive approaches to AI in education, this paper illustrates how professional learning through the African Engineering and Technology Network (AFRETEC) facilitated critical reflection and co-authorship. Rather than viewing AI as a neutral tool, the authors argue that dominant technological paradigms often reinforce colonial hierarchies, and advocate for a reflexive, values-driven approach to AI implementation. This paper contributes to the field by foregrounding a uniquely African, multilingual, and decolonial approach to AI in higher education, offering insights into the transformative potential of South–South–North collaboration and co-created knowledge. It highlights the sociocultural embeddedness of AI and calls for innovation that is ethically grounded and locally relevant.