Is hands-on learning still necessary in the age of AI? A thematic review.

INTRODUCTION: Generative artificial intelligence (AI) is increasingly entering design education and influencing how students engage in the design process. Existing studies have widely explored the potential of AI for creative production and improving design efficiency. Less attention has been paid to whether hands-on learning is still necessary in the age of AI. METHODS: This study adopted a thematic review approach to examine changes in the role of hands-on learning in AI-supported design education and its educational value. The review was based on 32 peer-reviewed studies retrieved from Web of Science and Scopus. RESULTS: The reviewed literature suggests that AI does not simply replace hands-on learning but changes how it may take place. Two interrelated forms can be distinguished: Embodied Hands-on, based on the body, tools, and materials, and Cognitive Hands-on, based on language, judgement, and human -AI iteration. Both share a cycle of action, feedback, reflection, and refinement, but their educational roles are not equivalent. The review also identifies concerns including the compression of learning processes, dependence on AI-generated outcomes, and an increasing emphasis on evaluation. DISCUSSION: This study proposes a process-oriented framework for understanding how Embodied Hands-on and Cognitive Hands-on may relate to creative thinking, design judgement, and student agency. The findings suggest that design literacy may increasingly involve critical evaluation, metacognitive regulation, and decision-making in human -AI collaboration.