@article{article_1717602, title={THE ROLE OF ARTIFICIAL INTELLIGENCE IN DESIGN: A STUDY ON THE REPRODUCTION OF PRESTIGIOUS UNIVERSITY LOGOS}, journal={Sanat ve Tasarım Dergisi}, volume={15}, pages={397–424}, year={2025}, DOI={10.20488/sanattasarim.1717602}, author={Dursun, Faruk}, keywords={Üretken yapay zeka, Yapay zeka destekli tasarım, Üniversite logoları, Görsel kimlik yeniden üretimi.}, abstract={This study aims to understand the capacity and limitations of AI tools in graphic design
processes. The research evaluates the potential impact of AI on corporate identity design
through the reproduction of logos of prestigious universities. The research analyzes in detail
the capacity of AI to interpret visual identities during logo reproduction. The research focuses
on examining the current capabilities of AI technology in the context of reproducing
university logos. In particular, the study seeks to answer the following questions:
- To what extent can AI tools accurately reproduce symbolically and aesthetically complex
visual identities?
- How are the visual elements of corporate identities interpreted and redesigned by AI?
The methodology of this research consists of three main phases: Data collection, reproduction
with AI tools and evaluation. In the data collection phase, the logos of the top
15 universities in The Times Higher Education World University Rankings 2023 list were
analyzed. However, 5 logos consisting only of typographic elements were excluded from the
scope of the study. Therefore, 10 university logos with visual and symbolic diversity (Oxford,
Harvard, Cambridge, Stanford, MIT, Princeton, Columbia, Chicago, Pennsylvania
and Johns Hopkins) were analyzed. The meaning and context of each university logo was
compiled based on the descriptions from their official websites. These descriptions detail the
logos’ history, design elements and symbolic meanings. The logos were reproduced using
10 different generative AI applications (Artguru, CoPilot, Design.ai, Gemini Advanced,
Genraft, Gettin.ai, Leonardo.ai, Chat GPT 4.0, Pixlr and Runwayml). These tools were
selected based on their visual production capabilities and user-friendly interfaces. For each
logo, descriptions taken from the official website of the relevant university were given as
“prompt” (input) to the artificial intelligence. Each logo produced by the AI tools was evaluated
according to the following criteria:
- Symbolism and Harmony of Meaning: The extent to which the logo generated by the AI
reflects the symbolic meaning of the original logo was analyzed.
- Design Accuracy: The similarity of the design elements was analyzed by comparing it with
the original logo.
- Detail and Composition: The accuracy of visual elements such as color palette, shapes and
general arrangement were evaluated.
As a result, in the context of detailed and aesthetic design; Leonardo.ai and Artguru offer
successful designs that reflect aesthetic harmony, richness of detail and academic identity
in Cambridge, Columbia and Harvard logos. Leonardo.ai stands out especially in the Columbia
and Harvard logos, while Artguru combines traditional elements with modern in
the Cambridge logo. In terms of preserving historical elements, RunwayML and Artguru
successfully modernize historical elements in the Cambridge, Columbia and Chicago logos,
reflecting academic identity. RunwayML in particular has managed to create a contemporary
feel while retaining traditional visual elements. Gencraft, CoPilot, Pixlr and Chat
GPT 4.0 produced effective outputs with minimal and modern designs in the Stanford and
MIT logos. Especially CoPilot offers a simple but remarkable solution in the MIT logo. Chat
GPT 4.0, on the other hand, displays a balanced modernity in the Columbia and Harvard
logos. Design.ai provides strong visual expressions in the Columbia and Harvard logos. The
Chicago logo is characterized by minimalism, while the Chicago logo is characterized by
artistic solutions, although some of the logos are complex. Gemini Advanced has successfully
applied modern design techniques to the Cambridge, Columbia and Harvard logos.
The Columbia logo stands out for its balanced design, even if it partially ignores traditional
elements. Pixlr and Getting.ai, on the other hand, failed to reflect the corporate identity
strongly enough in some logos (e.g. Pennsylvania and John Hopkins). Getting.ai tended to
over-simplify in the Stanford and John Hopkins logos, while Pixlr was weak with a lack of
detail in the Cambridge and Harvard logos.}, number={1}, publisher={Anadolu Üniversitesi}