Generative artificial intelligence as a design collaborator: A review of empirical studies


Ulusoy A. P., Nuhoğlu Kibar P.

International Visual Literacy Association 2026 Conference, 11 Mart 2026, (Yayınlanmadı)

  • Yayın Türü: Bildiri / Yayınlanmadı
  • Hacettepe Üniversitesi Adresli: Evet

Özet

The influence of generative artificial intelligence (GenAI) on learning processes has become a critical area of inquiry, particularly for its potential to support design practices by enabling the creation of visual content and functioning as a design collaborator that guides and inspires learners. In light of this potential, this study reviews twelve empirical studies in which learners engaged with GenAI to produce visually intensive outputs and examines how these experiences shaped their learning. 

The reviewed studies encompassed diverse tasks, including instructional design, art therapy visuals, logo design, argument and mind maps, digital advertisements, multimodal outputs such as video dramas, digital stories, and dance choreographies, as well as lesson plans and creative texts. Across these studies, the findings indicate that GenAI supported learners’ engagement, skill development, and creative production in diverse contexts. Learners often described GenAI tools as design collaborators, valuing their ability to provide frameworks, rapid feedback, and inspiration. Reported improvements included critical thinking, argumentation, creativity, empathy, language and communication skills, artistic and design tasks, multimodal production, as well as enhanced visual expression (through symbols and colors), knowledge structuring with visual representations (e.g., argument and mind maps), and the integration of text and visuals in multimodal products. At the same time, some studies noted limitations, reporting that outputs could lack originality and be repetitive, which underscored the continued importance of teacher support, particularly for conceptual clarification, qualitative evaluation, emotional depth, and social interaction.

Overall, the evidence suggests that AI-supported tasks can enrich both learning processes and outcomes when combined with teacher support as well as learners’ self-control and reflective, ethical use, and further research is needed to explore how learners can more effectively engage with AI as design collaborators in producing visually intensive outputs.