About this resource
This paper explores the intersection of artificial intelligence (AI), decolonisation, and indigenous knowledge systems within selected African tertiary libraries. It argues that current AI-powered information repositories often perpetuate Eurocentric epistemologies, marginalising African perspectives and reinforcing epistemic injustice in higher education. By proposing an Afrocentric framework, the study highlights how AI can be re-engineered to respect cultural diversity, integrate indigenous knowledge, and promote equitable information retrieval. Drawing on documentary analysis of scholarly articles, institutional reports, and policy frameworks from selected African tertiary institutions, the research identifies four critical dimensions requiring transformation: metadata epistemologies, training datasets, community collaboration, and ethical safeguards. The findings reveal that existing metadata standards cannot adequately represent oral traditions and community held knowledge; AI models are trained predominantly on Western datasets, with less than 1% reflecting African linguistic contexts; communities are excluded from decisions about how their knowledge is digitized and represented and ethical frameworks remain at the level of principle without enforceable mechanisms. The paper argues that decolonising algorithms is not only a technical necessity but also an ethical imperative for African higher education. An Afrocentric framework must operate across all four dimensions simultaneously, shifting from extractive data practices toward community data sovereignty and epistemic justice. The paper concludes that epistemic justice requires moving beyond procedural inclusion toward fundamental restructuring of AI systems to centre African knowledge systems, languages, and community governance structures. The framework proposed a path toward genuinely inclusive information repositories, but its realization depends on sustained institutional commitment, adequate resourcing, and genuine collaboration among institutions, communities, and library professionals.
Files and access
Resources are read online through DARE. Direct file links are not published; access is checked against DSpace when a document is opened. Authorised librarian workflows are unaffected.
Decolonising the Algorithm: An Afrocentric Framework for AI-Powered Information Repositories in Selected African Tertiary Libraries
Online reading · PDF document · rights governed by the repository
