Transparency is a Double-Edged Sword: Stakeholders’ Perception of GenAI Usage Disclosure in Kenyan Newsrooms

Generative artificial intelligence (GenAI) has been increasingly adopted globally in journalism, raising critical ethical considerations, especially regarding the transparency of its usage in news production. Existing research shows that disclosing AI authorship or involvement in news writing affects news audiences’ perceived trustworthiness and credibility, with most studies conducted in the Global North. However, current research has primarily focused on news audiences, overlooking other key stakeholders, such as journalists and policymakers, who also shape ethical norms and practices surrounding AI transparency in journalism. Furthermore, research on GenAI in journalism remains heavily focused on the Global North, leaving significant gaps in understanding how stakeholders in the Global South perceive and evaluate GenAI disclosure in journalistic content. In response, we conducted 94 in-depth interviews with six groups in Kenya—news audiences, journalists, policymakers, influencers, journalism educators, and information technology experts—to examine how stakeholders in Kenya perceive GenAI disclosure in journalistic content and the extent to which they trust news stories generated by AI or assisted by AI. Our analysis shows a “double-edged sword effect”: While most stakeholders advocated for transparency about GenAI use in journalism, many journalists expressed concern that such disclosure could undermine the perceived authenticity and credibility of news content. Such distrust stemmed from ethical concerns that overreliance on GenAI could lead to “less human” journalism and the spread of unverified or biased GenAI content. Stakeholders also expressed conditional trust that varies depending on the type of story, the level of GenAI use, and the nature of the content.