Diffusion, Logics, and Boundaries: How Journalists in Marginalized Media Markets Negotiate AI in Newsrooms

Literature on artificial intelligence (AI) and its role in journalism is expanding rapidly. Yet it remains disproportionately focused on Western contexts, ignoring how AI is perceived and negotiated in marginalized or semi-peripheral markets. This article focuses on how journalists in Albania, Croatia, Kosovo, North Macedonia, Romania, Slovenia, and Turkey perceive, apply, and challenge AI technologies. Drawing on theories of diffusion of innovations, institutional logics, and boundary work, this study analyzes the presentation of AI as both a technological innovation and a symbolic threat to professional identity. We conducted 10 focus-group sessions with 68 journalists and editors from diverse media outlets in seven different countries. Sessions were audio-recorded, transcribed, and thematically analyzed. Transcripts were independently coded by two researchers using AI-assisted coding in ATLAS.ti, alongside standard manual coding, with consensus-building to refine the process. The results show skewed and uneven AI implementation, motivated by efficiency requirements and not institutional strategy. Journalists present AI as a helping hand, noting concerns about fact-checking, editorial control, and professional autonomy. These strains indicate conflicting institutional logics, especially between technological efficiency and professional norms, and produce particular kinds of boundary work. Whereas defensive strategies underline the unreliability of human judgment, adaptive strategies also incorporate AI under editorial oversight. The article is relevant to the literature because it emphasizes ways in which structural limitations, institutional stresses, and professional ideals influence AI adoption outside of Western media systems.