How do DIGG guidelines for generative AI relate to the values of the Swedish Public Service?

In this essay i study how the Swedish Agency for the Digital Government`s (DIGG) national guidelines for generative Artificial Intelligence relate to the values of the Swedish Public Service. The purpose is to explore whether public services values are reflected in and in what way they are reflected or potentially challenged by the national guidelines. Two research questions guide the study: a. In what way are value base principles expressed in DIGG`s guidelines? b. What tensions can be identified between the guidelines and the principles of good governance? The study applies a qualitative policy analysis, using the Administrative Value Base - which establishes principles of good governance, as an interpretative/ framework for evaluating DIGG`s policy recommendations. The empirical material consists of DIGG`s policy recommendations, an interview with Karlstad municipality which is used for illustrative purposes only to highlight potential tensions that might appear in practice between established principles of good governance and guidelines. The interview should not be considered broadly applicable. The findings show that while principles such as legality, objectivity and transparency are echoed in the guidelines, tensions emerge when efficiency and digitalization are prioritized, raising questions about institutional legitimacy in public administration. The analysis is guided by sociological institutionalism which serves as a theoretical lens for interpretating the relationship between DIGG`s guidelines and the public service values. This perspective illustrates how traditional bureaucracy and the “rules of the game” is shaped when digitalization meets the normative principles of good governance.