”EN SKITSMART, DUM ASSISTENT”. En kvalitativ intervjustudie om hur yrkesverksamma inom kommunikation förstår och förhåller sig till generativa AI-verktyg

Abstract

This thesis focuses on the intersection of organizational communication and generative AI tools. It highlights the rapid evolution of generative AI tools, which has become more accessible to the general public, and discussed its significant societal impact, with specific attention to strategic communication within organizations. The purpose of the study was to investigate how communication professionals understand and relate to generative AI tools. The study aimed to elevate the understanding of the ideas and perceptions that communication professionals have about AI, focusing on what they consider to be the opportunities and limitations presented by the technology. This study reviews the historical development of media technologies, including AI, which have reshaped the media landscape and influenced organizational communication strategies. It discusses the evolution from traditional media to modern AI tools and the impact of these changes on strategic communication. The literature review noted a lack of comprehensive research on AI's role from a sender’s perspective, which this study aimed to address. This thesis utilized neo-institutional theory and the concept of imagined affordance. Neoinstitutional theory provided a perspective that deepened the understanding of the social and institutional dynamics influencing the valuation and use of artificial intelligence within organizations. This theory clarified how AI, as an institutional force, both shaped and was shaped by organizational norms and values. Imagined affordance explored how users perceived and interacted with technology, focusing on the collaboration between humans and media, where the user's perceptions and expectations of the technology are given more prominence. These theories provided a theoretical framework for the thesis. The methodological approach was a qualitative interview study involving nine semi structured interviews with communication professionals from both private companies and government agencies in Sweden. The choice of qualitative methods allowed for a detailed exploration of the complex perceptions and uses of AI within these fields. The participants included three individuals from government agencies and six from private companies, ensuring diverse insights into the institutional and organizational use and perception of AI. The study's results indicate that professionals in communication view generative AI tools as valuable resources, primarily for content creation, idea generation, and administrative tasks, with ChatGPT being particularly prominent. However, AI-generated content often requires significant human editing to meet professional standards. There are notable differences in how AI usage is regulated: government agencies have strict rules due to data security concerns, while private companies allow more individual discretion. Generative AI tools are valued for enhancingefficiency and creativity, but there are some skepticisms regarding their reliability and ethical implications. The study highlights that while AI can automate routine tasks, it cannot replace human-centered activities such as relationship-building. Skills of generative AI tools is seen as a competitive advantage, suggesting that professionals who quickly adapt to these technologies will thrive. Ultimately, the role of communicators is expected to evolve, with a shift towards directing AI-driven content creation rather than manual production. Generative AI tools are rapidly evolving. The study's findings should be viewed in the context of its timeframe (March-May 2024). The sample was uneven, affecting the second research question's results, but the conclusions remain valid. Private companies and government agencies regulate AI differently, reflecting varying priorities and responsibilities. Effective use of AI depends on the user's understanding and are influencing outcomes significantly. The study highlights the sociotechnical interaction, showing how technology and social processes shape each other. Future Research Suggestions: (1) Conduct a quantitative study on AI regulation in various organizations (2) Compare AI-generated and human-generated content for quality (3) Explore audience perceptions of AI-generated versus human-generated content