Between Opportunity And Governance: Discursive Constructions Of AI In Swedish Newspapers

Sammanfattning

This thesis investigates how artificial intelligence (AI) is discursively constructed in Swedish newspapers and how these constructions have evolved between 2020 and 2025. Drawing on qualitative critical discourse analysis (CDA) combined with the Social Construction of Technology (SCOT) framework, the study analyzes 41 articles from major Swedish news outlets. The aim is to examine how AI-enabled tools are discursively constructed, which social meanings and responsibilities are produced through these constructions, and how these representations develop over time. The analysis identifies four dominant discourses: AI as a productive and transformative infrastructure, as a disruptive force destabilizing institutions, as a source of risk and harm, and as a governance and democracy problem. These discourses do not replace one another; they coexist and build on one another, resulting in a layered media landscape. Over time, the discourses become more complex, showing competing interpretations related to efficiency, uncertainty, ethical concerns, and regulation. The findings demonstrate that multiple social groups shape AI discourse but remain dominated by institutional and expert actors, indicating an uneven distribution of interpretative power. At the same time, responsibility for adapting to AI is often placed on individuals. A key conclusion is the emergence of a paradox between governance and competitiveness: AI is discussed as both needing regulation and control and as essential for national development. This study highlights the central role news media might play in shaping public understandings of AI and raises broader questions about power, responsibility, and democratic participation in technological development.