ARTIFICIAL INTELLIGENCE FOR INCLUSIVE DEVELOPMENT IN WEST AFRICA:

Abstract

Although artificial intelligence has the potential to drive positive change and spark innovative product growth and development, such advances carry enormous risks of reinforcing and expanding existing inequalities amidst the increasing adoption of AI systems in various segments of public life in West Africa (Crawford, 2021; Eubanks, 2018). This paper advocates a comprehensive and cross-disciplinary analysis of inclusive development in the framework of AI, to explore three interrelated areas, namely public service delivery, maternal health care and preservation of indigenous languages. These areas, usually studied independently, have common structural weaknesses: governance gaps, lack of data infrastructure, digital exclusion, and a centralisation of knowledge and decision-making with AI in institutions from the Global North. Based on an intercountry comparison (Nigeria, Ghana and Senegal) using a multi-case study design that integrates data from all three countries, this paper aims to analyse how AI is being adopted, controlled and experienced by its users across contexts, with a focus on identifying shared trends and distinguishing similarities and differences in ethical, institutional and distributional implications. There is a steady momentum of tension between technocratic optimism in seeing AI as a development solution and critical concern that it could reproduce some of the structural inequalities, governance failures and epistemic marginalisation that keep occurring (Couldry & Mejias, 2019; Joshi et al., 2020). This paper presents an integrative framework that integrates several issues related to AI's impact on West Africa's development, including equity-oriented design, democratic governance, epistemic justice of West African peoples' knowledge, institutional capacity-building, and regional solidarity based on cooperation among West African countries for shared AI infrastructure, as well as data governance and regulatory harmonisation