Decolonising Generative Artificial Intelligence in African Higher Education: Toward Ubuntu-Ẹ̀kỌ́, Augmented Collective Wisdom and Digital Sovereignty

Abstract

The rapid diffusion of generative artificial intelligence (GenAI) into higher education presents both significant opportunities and critical challenges for institutions in Africa and the wider Global South. While GenAI can expand access to information, support teaching and learning, and strengthen research productivity, its deployment also raises questions concerning whose knowledge is represented, which languages are privileged, how credibility is constructed, and who exercises control over the infrastructures and data on which artificial intelligence depends. This article critically examined the implications of these challenges through the lenses of decolonial theory, epistemic justice, African philosophical thought, and critical AI studies. Rather than treating artificial intelligence as a culturally neutral technological artefact, the article conceptualises AI as a sociotechnical system whose knowledge practices are shaped by historical, economic, linguistic and epistemological relations of power. Drawing on a critical interpretive synthesis of scholarly literature, policy documents and African philosophical perspectives, the article developed the Ubuntu-Ẹ̀kọ́ Framework as an original conceptual contribution. The framework brings together relationality, ethical character, contextual responsiveness, critical consciousness and communal capability. Building on this framework, the article proposes Augmented Collective Wisdom (ACW) as a reorientation of AI from a technology primarily designed to automate or individualise knowledge work toward a sociotechnical resource that supports collective learning, human judgement, cultural integrity and community wellbeing. The article further advanced Digital Sovereignty as a strategic alternative to a narrow discourse of AI readiness. It concluded that decolonising AI in higher education does not require rejection of technological innovation; rather, it requires institutions to develop the epistemic, cultural, linguistic, infrastructural and governance capacities necessary to determine how AI is designed, interpreted, deployed and evaluated within their own contexts.