“AI Will Not Replace Journalists, but a Journalist Who Uses AI Will”: AI Integration and the Changing Boundaries of Journalistic Work in Public Service Media

Sammanfattning

This study examines how AI is integrated into journalistic work within Swedish Public Service Media (PSM), focusing on how journalists and key organisational actors at Sveriges Radio experience and negotiate the growing presence of AI in newsroom practice. The study investigates how AI tools are incorporated into journalistic workflows, how journalists understand and negotiate the role and boundaries of AI use in relation to their perceptions of journalistic and PSM values, and how they describe the opportunities, challenges and educational needs associated with increasing AI integration. The study draws on the Structurational Model of Technology, Institutional Theory and perspectives on Public Service Media Legitimacy in order to analyse how AI is negotiated and legitimised within an institutional journalistic context. Methodologically, it uses a qualitative case study design based on semi-structured interviews with journalists, an editor-in-chief and an AI strategist at Sveriges Radio, with the empirical material analysed through thematic analysis. The findings suggest that AI is primarily legitimised when supporting routine and time-consuming tasks such as transcription, translation, summarisation, research and data processing, allowing journalists to focus more on editorial judgement and creative work. AI is also perceived as expanding journalistic practice by enabling new forms of analysis and problem-solving. However, it becomes contested when approaching tasks linked to authorship, accountability, creativity and human interaction. The study shows that AI integration generates tensions around competence, skill development, organisational expectations and journalistic identity, particularly within PSM, where legitimacy is tied to trust, transparency, editorial independence and democratic responsibility. It contributes to research on AI in journalism by framing adoption as an ongoing sociotechnical, institutional and professional negotiation, highlighting the need for situated AI literacy, continuous learning and critical human oversight.