Generative AI and Psychological Need Satisfaction and Need Frustration at Work: A Self-Determination Theory Perspective

Abstract

Generative AI (GenAI) is increasingly integrated into knowledge work, yet its consequences for workers’ psychological needs remain unclear. Drawing on Self-Determination Theory, this study aims to examine how knowledge workers integrate GenAI into their work, how this reshapes their task, knowledge and social characteristics, and how these changes are experienced in relation to autonomy, competence and relatedness. Based on twelve semi-structured interviews with Dutch knowledge workers, thematic analysis resulted in five themes. Participants used GenAI selectively, setting clear boundaries around when and how the tool could support their work. GenAI often amplified professional capability by making work faster, improving output quality, expanding task scope and supporting learning. At the same time, convenience related to the use of GenAI could reduce independent reasoning and increase reliance, a tension captured as the convenience trap. Autonomy was supported through enhanced freedom and flexibility but frustrated when AI-assisted output felt less self-authored. Relatedness was redistributed rather than replaced, as everyday practical support partly shifted from one’s colleagues to GenAI, while deeper collaboration remained tied to human colleagues. Overall, the findings suggest that GenAI's psychological effects are ambivalent and depend on how the tool becomes embedded in everyday work practices.