Relationality and Data Justice for Trustworthy AI Practices in Africa

Abstract The aim of this chapter is to unpack what is needed to ensure the trustworthiness of AI practices in Africa through the lens of social justice considerations. The chapter is a call for a strategy for building a sustainable equitable AI ecosystem in Africa supported through trustworthy AI practices focused on public and communal benefit rather than enriching Big Tech companies on the other side of the world. I introduce the notion of ‘AI justice’, which is justice for every inhabitant of the African continent who engages with AI technology at any stage of its lifecycle, and which is a notion embedded in a relational ethic and emerging from a combination of data and design justice approaches as elements of social justice in the domain of AI. In this sense, an AI practice will be trustworthy when it protects the rights and benefits of the communities whose data it uses. In this sense I will speak of trustworthy AI practices ‘serving’ communities. This chapter is a challenge to all AI actors in Africa—the researchers, designers, developers, deployers, and users—to claim their collective ownership of the domain of AI and stand together to build social resilience against the Big Tech business model that drives algorithmic colonisation.