"This study demonstrates impressively how widely AI is already used in South African newsrooms – often, however, without systematic training, without editorial guidelines, and without tools adapted to African languages and contexts. The interviews make it clear that journalists stand between curiosity and caution. They recognise the potential of AI for research, verification, data analysis, or translation, yet they are equally aware of risks such as algorithmic bias, hallucinations, the erosion of journalistic skills, and possible loss of trust among audiences. Thus, the central question is less whether AI should be used in newsrooms, but how: responsibly, transparently, critically, and always in service of journalistic integrity. The study shows that many newsrooms still lack the necessary foundations – training, internal rules, structures, awareness of risks, and technological solutions that are not based solely on data from the Global North. At the same time, the willingness to use AI meaningfully is clearly visible, especially where it strengthens journalistic capacity rather than replacing it." (Foreword)