Ontological disruption or data colonialism?Institutionalising Kibagenge for artificial intelligence integration in Kenyan Higher Education
Year: 2020
Type: article
Abstract: The introduction of Artificial Intelligence (AI) into Kenyan Higher Education can be described as the technological imperative to preserve Indigenous Knowledge Systems (IKS). However, this paper questions this integration as a possible site of data colonialism, a condition in which indigenous epistemologies are packaged as discrete, alienable information packs to be trained by Western algorithms. This study utilised a Critical Policy Discourse Analysis (CPDA) of a corpus of 15 AI policy documents spanning global mandates, continental strategies, and Kenyan national laws. An analysis of public and private universities’ AI guidelines was also carried out to identify a systemic failure. The findings reveal that the ontological erasure of the communal, relational character of IKS is a consequence of institutions’ uncritical adherence to individualistic Western legal frameworks. To fill this gap, the study proposes a new Kipsigis framework, Kibagenge. This paper proposes that to adequately decolonise AI in institutions, extractive governance models must be replaced with Indigenous Data Trusts (IDTs) and Kokwet ethical protocols, which enforce collective consent. The paper concludes that Kibagenge is not merely a cultural critique but a required methodological and policy construct for reclaiming epistemic sovereignty in Kenyan higher education. The paper’s contribution to scholarship is threefold: (i) it formalises Kibagenge as a decolonial framework for AI and data governance; (ii) it operationalises the Agenge / Tiliet / Chamyet triad into a five-tier policy diagnostic (global → continental → national → public university → private university); and (iii) it proposes concrete institutional instruments, Indigenous Data Trusts, the Kokwet ethical protocol, and Tiliet-based procurement, for Kenyan universities.
Author: Amon Kipyegon Kirui
Institution: University of Venda
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
Funders: Universität zu Köln, African Union
OpenAlex ID: W7214911134