Artificial Intelligence Practices in Everyday News Production: The Case of South Africa’s Mainstream Newsrooms

This article explores artificial intelligence (AI) uptake in selected South African mainstream newsrooms. It seeks to determine the extent to which AI has been adopted and how journalists and editors perceive its appropriation in newsmaking practices. To address these two broad aims, the study used in-depth interviews with journalists and editors. Our findings suggest a slow, varied but methodical uptake of AI practices in South Africa's mainstream newsrooms. We deduced three uses of AI in these newsrooms. The first is what we call the holistic appropriation of AI. The second one is the exclusively technological appropriation of AI, and the last one is the task-specific appropriation of AI. This varied uptake of AI is taking place against a deep-seated skepticism with this technology. The skepticism is driven by fear of job losses, the costs of adopting AI, limited training, ethical issues around AI and its efficacy in the democratic process. On this last point, South African journalists question whether AI can be beneficial to the sustenance of a post-apartheid democratic society. Our argument, therefore, is that the optimism about AI in newsrooms that some researchers find in Europe and US newsrooms cannot be transferred to newsrooms in South Africa.