Mellan effektivisering ochmänskligt omdöme : En kvalitativ studie om generativ AI i offentlig sektor

Sammanfattning

Generative AI (GenAI) is becoming increasingly common in the public sector and isbeing used in a growing number of work processes related to decision making. At thesame time, the technology raises questions regarding accountability, transparency, andhuman oversight. The purpose of this study is to examine how public sector employeesperceive and use GenAI in work processes related to decision making. The studyadopted a qualitative research approach and was based on nine semistructuredinterviews with employees from different public sector organizations. The empiricalmaterial was analysed using thematic analysis.The findings show that GenAI is primarily used for information retrieval, informationprocessing, text generation, analysis, and the preparation for decision support materials.Respondents perceive the technology as contributing to time savings, reduced cognitiveworkload, and increased autonomy in their work. At the same time, the findings revealthat AI-generated content requires continuous review and verification due to the risk ofinaccurate or misleading information. Furthermore, the results indicate that timepressure, high workloads, and a lack of clear guidelines influence the use of GenAI. Thestudy suggests that GenAI has significant potential to support work processes related todecision making within the public sector. However, its effective use depends on humanoversight, professional judgement, and clear organizational conditions. The studycontributes empirical knowledge regarding how GenAI is used and perceived within thepublic sector. Keywords: Generative AI (GenAI), public sector, work processes related to decisionmaking, public-sector employees, human oversight