This master thesis examines how the integration of artificial intelligence (AI) systems into organizational workflows reshapes decision structures and governance in hybrid and exception-based settings. The study addresses a gap in the literature concerning how AI becomes embedded in organizational decision-making processes, how it reshapes information processing, decision responsibility, and oversight are maintained in AI-supported environments.

This research adopts an abductive qualitative design and draws on 10 semi-structured interviews with professionals in strategic and decision-related roles across organizations from different sectors and national contexts. The data were analyzed using thematic analysis.

The study contributes to research on organizational decision-making and AI governance by showing that decision-making is increasingly shaped by AI systems that filter, prioritize and structure information before human interpretation, thereby influencing how decisions are formed. In addition, AI affects role distribution in decision-making processes by influencing when and how human actors participate across different stages of decision-making, while highlighting the emergence of ongoing monitoring of AI systems as part of governance practices. Governance practices are embedded in system design and enacted through escalation routines, threshold-based interventions, and user competence in interpreting AI outputs.