The Applications and Roles of Generative Artificial Intelligence in School-Aged Student Creativity Development: A Systematic Review

Abstract

Generative artificial intelligence (GenAI) has rapidly entered creativity-oriented education since 2022, yet evidence on its applications and roles in school-aged student creativity development (SCD) remains fragmented. This systematic review addresses two questions: (1) What architectures, tools, and modalities of GenAI support SCD? and (2) What facilitating and constraining roles do GenAI primarily play across creativity dimensions, and what inter-role dynamics emerge among them? Following PRISMA 2020, we searched five databases and two supplementary sources, identifying 57 empirical studies published between 2022 and 2026. Descriptive analysis mapped GenAI architectures, modalities and tools, while inductive–deductive coding synthesised GenAI roles using Rhodes’s 4 Ps creativity framework. Results revealed that GenAI tools built on transformer-based large language models and diffusion models and operating through text-to-text and text-to-image modalities dominated applications. Regarding GenAI’s roles in SCD, we identified 20 distinct roles, 19 of which were distributed across 4 Ps, and revealed three dynamic inter-role patterns. This study adds a new insight to Rhodes’s 4 Ps by incorporating the role of Co-Creator as a fifth P, Partnership, to capture the proactive, dialogic, and distributed nature of human-machine co-creation. Implications are discussed for “role-aware” tool selection, pedagogy, and future research on underrepresented modalities and partnership dynamics.