Exploring Trustworthy AI in Nigeria: A Focus on Safety in Road Traffic

Abstract

Safety is paramount in AI systems’ adoption and trust. Much of the work on trustworthy AI principles and practices are primarily from Western contexts, despite the often global deployment of AI-enabled systems. The lacking representation of African contexts and voices in AI safety discourse may lead to disproportionate harm from AI systems in Africa. In this chapter, we contribute an African perspective to AI safety by exploring safety in Nigerian road traffic—an environment where AI-enabled systems designed in Western contexts are gaining traction. With the view that safety is systemic, we explored (1) AI safety definitions and (2) practical safety in Nigerian road traffic by highlighting some of the safety impacts of automobility and its regulation in Nigeria. Drawing on these, we identified sociotechnical and environmental factors that are critical for safe AI development and adoption in Nigeria. We conclude the chapter by discussing the lessons and recommended actions for ensuring AI safety in Nigeria and other African countries. They include systemically prioritising Africa’s and Africans’ safety in AI systems through research and stakeholder engagements, local and global empowerment of African researchers, respecting African values and contributions to global AI discourse, and creating spaces for communal contributions to AI safety through public engagements and education. Show more