The integration of Generative AI in the design search process

Context: The integration of Generative AI in educational practices has been rapidly evolving, particularly within design programs. Institutions like Singapore Polytechnic have begun leveraging generative AI tools, such as MidJourney and DALL·E, reshaping how students approach ideation, research, and creativity in a technology-driven environment. Aims: This study aims to explore the impact of a generative AI-assisted search workflow on design students' research and ideation processes. Specifically, it investigates how this integration affects students’ design vocabulary, variation in visual references, ability to justify visual choices, and overall design quality, guided by Puentedura’s SAMR model, particularly at the Modification level. Methods: The case study involved 45 second-year Visual Communication and Motion Design students. Data were collected through content analysis of submitted work, student and lecturer interviews, and assessment rubrics spanning five design briefs. A comparative analysis was made between students' abilities before and after exposure to the generative AI search methodology, with a focus on four specific research questions. Findings: Results indicated that higher readiness students experienced a slight expansion in their design vocabulary and variation in visual references. However, both high- and low-readiness groups struggled to adopt a broader range of visual references, revealing a tendency to rely on traditional search methods alongside AI. Students' abilities to justify their design choices and the overall quality of final designs showed minimal improvement, with many relying on AI-generated content without critical assessment or integration into their creative processes. Implications: The findings suggest that while generative AI can enhance ideation speed, it often does not deepen critical design reasoning or vocabulary development. The study highlights the need for structured frameworks that promote deeper engagement with AI outputs and a more diverse approach to research, suggesting that exposure to technology should be accompanied by explicit instruction in design principles and critical evaluation techniques to ensure meaningful learning outcomes.

Keywords (free text)

Generative AI, design education, ideation, search strategies, SAMR model

Permanent link to cite this item

https://hdl.handle.net/20.500.14069/1337
https://doi.org/10.71779/793

Collections

DI-IDEA AI for Education (AI4E) Competition 2025