International Journal of
Information and Education Technology

Abstract

Integrating Generative Artificial Intelligence (GAI) into Project-Based Learning (PjBL) provides new approaches to foster students’ Creative Thinking (CT) in solving real-world design problems. However, empirical evidence on its impact, particularly on CT, student engagement, and Artificial Intelligence (AI) literacy, remains limited. This study investigates the effects of a GAI-supported PjBL model (PjGAI), in which Kimi and MasterGoAI were systematically embedded within a staged project-based design workflow. A total of 180 undergraduate students were randomly assigned to three groups in an experimental design. All groups used the same conventional digital design software for User Interface (UI) production. The PjGAI group additionally used the designated GAI tools, the PjBL group followed the same project-based workflow without GAI support, and the Traditional Teaching Model (TTM) group followed a lecture-based, teacher-led workflow without GAI support. Multivariate Analysis of Variance (MANOVA) results indicated that the PjGAI group achieved significantly higher CT scores than the TTM group and reported higher student engagement than both comparison groups, although the CT difference between the PjGAI and PjBL groups was not statistically significant. Within the PjGAI group, paired-sample t-tests showed a significant increase in AI literacy from pre-test to post-test. These findings suggest that embedding GAI tools into a structured PjBL process can strengthen student engagement and may support creative exploration and AI literacy development in design education.