== 17.06.2025 - 13:00 - 14:00 (local Lübeck time, CEST): Online via WebEx only.
Epistemic justice: A case for inclusion of indigenous African perspectives in the Sustainable AI discourse By Jahaziel Osei Mensah (University of Ghana)
Abstract: Epistemic justice requires the recognition and inclusion of the often-marginalized voices in artificial intelligence (AI) ethics. Eke et al (2023) find that African perspectives of sustainable AI are sorely lacking, yet Africa incurs the brunt of the repercussions of AI use by the Global North. This has resulted in the adoption of Western ethical frameworks on sustainable AI in Africa, and further exacerbated the environmental cost of the life cycle of AI in Africa (extraction of minerals, carbon emissions through training of AI models, and E-waste disposal). African tradition does not merely conceive of the environment as a means to an end, but also grasps the environment as an end in itself. The environment is a habitat of God, ancestors, and deities, and therefore human beings are to treat it with respect. Therefore, an African-centred framework on sustainable AI will prioritize environmental sustainability. To have robust AI regulations in Africa that are based on African communitarian values, there is a need to incorporate African communitarian epistemologies. In this presentation, I lay bare the epistemic gap in the sustainable AI debate. I begin by discussing the lack of representation of African views, and further discuss the concerns in the sustainable AI debate. Lastly, on the need for epistemic justice, I will propose ways through which African ethical perspectives can contribute to the goal of sustainable AI (van Wynsberghe, 2021).
