Decolonizing artificial intelligence ethics : Ubuntu as a philosophical paradigm for AI governance in Africa

Abstract

This dissertation examines the relevance of the African philosophy of Ubuntu to the ethical governance of artificial intelligence in African contexts. It analyses dominant global frameworks for AI ethics and governance and identifies conceptual and normative limitations in their engagement with African social values, communal worldviews, and historical experiences. The study adopts a qualitative, normative, and interpretive approach, drawing on African philosophical literature, international and regional policy documents, and expert interviews to assess how ethical assumptions are embedded in existing AI governance models. The analysis conceptualises Ubuntu as a relational ethical framework centred on personhood, collective responsibility, and interdependence, and examines how these principles challenge prevailing individual-centred approaches to AI ethics. It explores the implications of Ubuntu-informed ethics for data governance, accountability, and institutional design, with particular attention to policy and legal frameworks at national, regional, and continental levels. By positioning African philosophical thought as a source of normative reasoning rather than contextual adaptation, the dissertation contributes to discussions on ethical pluralism and legitimacy in AI governance. It concludes by identifying governance considerations relevant to African institutions and reflecting on the broader implications of Ubuntu-informed ethics for global debates on the responsible development and use of artificial intelligence.