When individuals use generative AI, the diversity of ideas decreases sharply. When entire teams rely on the same AI tools, the convergence compounds. Most research to date has focused on enhancing individuals’ creative abilities through use of generative AI. However, no prior studies have examined how distributing different AI tools across team roles affects creative collaboration. This research asks whether providing different AI capabilities to each role within a team can help counteract this trend toward homogenisation. This paper introduces the “AI as Fifth Team Member Model” for organising student design teams and their respective AI tools. The model defines four different roles: Research Coordinator (using ChatGPT or Claude), Visual Design Specialist (Midjourney or DALL-E), Space Planning Expert (Planner 5D or Homestyler), and Presentation Coordinator (who uses Canva AI or D5 Render). Each tool mediates creativity in a different way, through text-based analysis, image generation, 3D modelling and layout design, which forces heterogeneity into the team’s creative process. Each role also produces a different kind of output, and combining them requires human deliberation. That is where the diversity comes from. A pilot of the model took place in two workshops at a Scandinavian university of applied sciences: a four-day bachelor-level workshop and a two-day master-level workshop. The pilot results indicated that four days were not enough for bachelor-level design teams to form a cohesive unit, and that daily synchronisation protocols (i.e., morning sync, midday integration, and end-of-day synthesis) are essential for facilitating cross-role coordination. The model is currently being used in an educational context, but the way tasks are divided among roles is similar to how many professional design studios now divide tasks among specialists. Future studies should assess larger groups and measure formally the increased diversity in outputs, to verify these preliminary findings.