From decision support to continuous improvement: GenAI appropriation, performance and self-assessment calibration

Purpose This study investigates how different modes of generative artificial intelligence (GenAI) appropriation are associated with two complementary dimensions of continuous improvement in decision-making. The first is decision-making performance, which reflects the cognitive ability to make effective choices; the second is pre-feedback self-assessment calibration, which captures the metacognitive ability to align perceived decision quality with objective outcomes across iterative decision-making cycles. Design/methodology/approach The study involved 36 managers from manufacturing companies enrolled in lean six sigma green belt courses. To assess decision-making performance over time, a multi-round business simulation game designed to replicate complex decision-making scenarios was employed. At the end of each simulation round, participants completed a questionnaire measuring their ability to self-assess the quality of their decisions. Semi-structured interviews were also conducted to examine how participants interacted with GenAI. Findings The results indicate that different modes of GenAI appropriation are associated with distinct trajectories of decision-making performance and self-assessment calibration. Reflective use of the technology was associated with higher decision-making performance and greater improvement in self-assessment calibration. Passive use, characterised by a high level of delegation of decision-making to GenAI, was associated with more volatile performance trajectories and poorer calibration. Progressively reflective use was associated with intermediate but improving trajectories over time. Originality/value This study introduces a longitudinal perspective on GenAI-supported decision-making by jointly examining decision-making performance and pre-feedback self-assessment calibration across repeated decision cycles. By relating different appropriation modes to cognitive and metacognitive trajectories, this study extends research beyond approaches focused on isolated performance outcomes and provides a process-oriented view of decision-making improvement over time.