The politics of translation in African-language generative AI: opportunities and challenges for critical engagement, education and resistance

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

This article explores the opportunities and challenges of knowledge exchange in East Africa around the societal implications of generative AI. These technologies’ capacity to work with and in certain African indigenous languages is increasing rapidly and generative AI is quickly becoming part of the region’s digital platform landscape – albeit in an unpredictable and experimental fashion. Drawing from the authors’ practice-based research in Somalia and their work on Somali-language educational and public engagement resources regarding generative AI, the article argues that translation is a crucial but ambiguous space through which critical debate and resistance to dominant tech narratives can be enabled. This is because of the centrality of language to the (in)accessibility of debates about AI, as well as the emerging impacts of AI on language and translation practices themselves. Building on previous work on the power and problematic aspects of predictive text algorithms that function in East African languages, the article emphasises the importance of evaluating the implications of rapid global AI developments from Eastern African vantage points, foregrounding historical inequalities around language and ongoing (data) extractivism.