Abstract
This study examines how AI image generation systems perpetuate digital orientalism through systematic cultural representation patterns that increasingly shape global digital markets and international development discourse. Using visual social semiotics methodology, we conducted comparative analysis of Midjourney and ChatGPT platforms across nine culturally specific prompts examining religious imagery, contemporary social dynamics, and power structures in Indian cultural contexts. Our findings reveal that both platforms consistently reproduce colonial-era representational frameworks despite culturally sensitive prompting and advanced technical capabilities. The analysis identifies three systematic patterns: temporal displacement strategies that acknowledge modernity while containing it within orientalist categories, institutional validation requirements that privilege formal over community-based authority, and aesthetic transformation that converts authentic cultural practices into consumable visual narratives for global markets. Platform differences emerge through Midjourney's artistic stylization versus ChatGPT's cultural essentialism, yet both privilege Western aesthetic expectations over authentic representation. These patterns demonstrate that digital orientalism operates through accumulated training data reflecting centuries of colonial visual documentation rather than explicit programming choices. The study contributes theoretical insights into how digital technologies reproduce colonial power relations while providing practical implications for developing culturally conscious AI systems. This analysis has urgent implications for global digital markets, showing that a fundamental transformation of technological development is needed to prioritize community expertise and dismantle colonial knowledge hierarchies.
Date Received
June 3, 2025
Date Revised
November 4, 2025
Date Accepted
November 14, 2025
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Varghese, Jinu K. and Rani, Padma
(2025)
"From Prompt to Prejudice: Visual Evidence from AI Image Generation Platforms,"
Markets, Globalization & Development Review:
Vol. 10:
No.
2, Article 4.
DOI: 10.23860/MGDR-2025-10-02-04
Available at:
https://digitalcommons.uri.edu/mgdr/vol10/iss2/4
Included in
Asian Studies Commons, Communication Technology and New Media Commons, Critical and Cultural Studies Commons, Digital Humanities Commons, E-Commerce Commons, Management Information Systems Commons, Technology and Innovation Commons
Author Bio
Jinu K Varghese is an AI-Human Interaction Research Fellow at Manipal Institute of Communication. Currently pursuing his PhD and serving as a Research Fellow at Manipal Academy of Higher Education, he brings over a decade of academic experience, including his role as former Head of Department (Multimedia) at St. Joseph College of Communication. He holds a master’s in digital media from The Catholic University of Korea and a Diploma in Film Appreciation from the Film and Television Institute of India, Pune. His research focuses on the intersection of artificial intelligence and visual communication, exploring how technology can enhance visual storytelling and bridge cultural divides.
Padma Rani is Professor & Director of Manipal Institute of Communication. She has over 15 years of teaching experience at the postgraduate level. Her areas of interest are gender, ICTs, media and communication policy, and human rights. Dr Padma Rani is also the Coordinator of the Media Research Centre, Manipal Institute of Communication. She has conducted research studies for the United Nations Population Fund and the National Commission for Women, New Delhi. She is the editor of Global Media Journal (India edition). Her recent publications include Widening the Wedge: Digital Inequalities and social media in India, Mainstream Media’s Framing of #Metoo Campaign in India, and Health from the perspectives of Indian rural populace: urgent need to mainstream mHealth with the emergence of covid-19.