Generative AI as a Design Collaborator: Redefining Authorship, Creativity, and Pedagogy in Contemporary Design Practice

Abstract

What happens to design authorship when the machine can generate a hundred variations of a facade before the architect has finished their first sketch? This question, which might have seemed hypothetical even five years ago, is now routine in architectural and graphic design studios worldwide. This paper examines how generative AI is unsettling three interlocking foundations of design practice: the attribution of creative authorship, the cognitive basis of design creativity, and the pedagogical conditions under which design judgment is formed. Rather than treating these as separate problems, the discussion tries to show how they reinforce each other, and why responding to any one of them in isolation is likely to produce incomplete and potentially counterproductive solutions. The paper draws on a range of recent empirical and theoretical work, including jury-based studio research and computational performance studies, and situates current debates within the longer history of design's relationship with disruptive technologies. The argument is not that generative AI is simply good or bad for design. It is that the profession is, at this moment, making choices whose consequences will be difficult to reverse, and that making them thoughtfully requires a clearer sense of what design has always been for, not just what it currently does.