How Does Artificial Intelligence Shape Creative Thinking Processes in Educational Contexts? A Systematic Review of Cognitive Mechanisms

Although artificial intelligence tools increasingly shape creative thinking in education, existing research focused more on outcome effects than on cognitive processes. This systematic review examined the cognitive mechanisms, process-phase transformations, and student strategies characterizing AI-integrated creative thinking. Guided by PRISMA 2020 (Page et al., 2021), 49 peer-reviewed empirical studies published between 2022 and 2025 were selected from Scopus and Web of Science. A process-evidence criterion framework ensured the inclusion of studies reporting process-level data beyond outcome-only measures. Methodological quality was assessed using the MMAT, and data were analyzed through content analysis and thematic synthesis. The most striking finding revealed that AI-mediated creative thinking is shaped by how learners engage with AI outputs. Across many studies, the same AI use was associated with both facilitative and inhibitory cognitive mechanisms within the same learning activity. This pattern suggests that the impact of AI depends not simply on using AI for part of the thinking process, but on whether learners engage with AI outputs critically or accept them uncritically. Accordingly, the central challenge is no longer simply whether AI can enhance creativity, but how educators can design learning environments that harness AI productively while preserving and developing the deeper cognitive processes that underlie human creative thinking.