This study investigates the complex interplay between organizationally prescribed perfectionism (OPP), job insecurity, counterproductive work behavior (CWB), and self-efficacy in learning artificial intelligence (AI) in the context of modern organizations. Based on several theories, the current research suggests and tests a moderated mediation model. Using a three-wave time-lagged design with data collected from 412 workers across various South Korean corporations, we examine how OPP influences CWB both directly and indirectly through job insecurity, and how self-efficacy in AI learning moderates the OPP–job insecurity link. Our results show that OPP is positively linked to CWB, and this association is partially mediated by job insecurity. Moreover, AI learning self-efficacy functions as a moderator in the OPP–job insecurity link, such that the positive link is weaker for members with higher levels of AI learning self-efficacy. These findings extend our understanding of perfectionism in organizational settings and highlight the role of technological self-efficacy in mitigating the negative impacts of perfectionist cultures. This research may contribute to the literature on perfectionism, CWB, and technological adaptation at work, and has important implications for managing high-performance cultures in the period of rapid technological advancement.