YAPAY ZEKÂ TARAFINDAN ÜRETİLEN GÖRSELLERDE BASMAKALIPLAR: TÜRKİYE TEMSİLİ ÜZERİNE BİR İNCELEME

Abstract

Yapay zekâ teknolojisinin her geçen gün artan gelişimi ve yoğun kullanımı beraberinde toplumsal ve kültürel boyutta endişe yaratan sorunları da getirmektedir. Bu sorunlardan birisi olarak görülen basmakalıplaşmış yargıların yeniden üretimi, özellikle görsel üreten yapay zekâ araçlarında kendini göstermektedir. Basmakalıplaşmış yargıların yeniden üretimi, negatif ve pozitif yönlü olabilecek ön yargıların sürdürülmesine sebep olmakta ve küreselleşmiş dünyanın aşılamayan düşünsel engellerini pekiştirmektedir. Bu çalışmanın da temel kaygısını bahsi geçen durum oluşturduğu için görsel üreten yapay zekâ araçlarına odaklanılmıştır. Üretken yapay zekâ araçlarının perspektifinden Türkiye ve Türkiyeli temsilini anlayabilmek ve bu temsillerde yer alan basmakalıp yargıları saptamak çalışmanın temel amacını oluşturmaktadır. ChatGPT, Copilot ve Gemini yapay zekâ araçlarının örneklem olarak seçildiği çalışmada, karma yöntem üzerinden araştırma tasarlanmıştır. Üretilen görsellerin nitel ve nicel içerik analizi ile detaylandırılması sonucunda Türkiye’ye özgü farklılıklarda fiziksel özelliklerin öne çıktığı bunun da kadın cinsiyetinde başörtüsü, erkek cinsiyetinde sakal ile bıyık şeklinde kendini gösterdiği anlaşılmıştır. Türkiyeli temsilini yansıtan bazı görsellerde sunulan geleneksel kıyafet kullanımı ise Oryantalist bakış açısını örneklendirebilen türden olmuştur.

Keywords

• Yapay Zekâ
• Basmakalıp
• Üretken Yapay Zekâ
• Ülke Temsili
• Karma Yöntem

STEREOTYPES IN AI-GENERATED IMAGES: AN ANALYSIS OF THE REPRESENTATION OF TÜRKİYE

Abstract

With the rapid advancement and increasing use of artificial intelligence (AI) technologies, concerns with significant societal and cultural implications have emerged.. The reproduction of stereotyped judgments, which is seen as one of these problems, especially manifests itself in image-generating AI tools. The reproduction of stereotyped judgments causes the continuation of prejudices that can be negative and positive and reinforces the insurmountable intellectual obstacles of the globalized world. Since the aforementioned situation constitutes the main concern of this study, the focus is on artificial intelligence tools that produce images. The main purpose of the study is to understand the representation of Turkey and Turkish identity from the perspective of generative artificial intelligence tools and to identify the stereotypes in these representations. The study adopts a mixed-methods approach and includes ChatGPT, Copilot, and Gemini as its sample AI platforms. As a result of the detailed analysis of the generated images with qualitative and quantitative content analysis, it was understood that physical characteristics stand out in the differences specific to Turkey, and this manifests itself in the form of headscarves for women and beards and moustaches for men. The use of traditional clothing in some of the visuals reflecting the Turkish identity exhibit characteristics of an Orientalist gaze.

Keywords

• Artificial Intelligence
• Stereotyped
• Generative Artificial Intelligence
• Representation of Country
• Mixed Method

References

1. Abbas, A. (2025, January 23), Western bias in AI: Why global perspectives are missing. Unite AI. https://www.unite.ai/western-bias-in-ai-why-global-perspectives-are-missing/
2. Agbo, K. (2024, June 24). Africans' contributions to AI can reduce bias. THISDAYLIVE. https://www.thisdaylive.com/index.php/2024/06/24/africans-contributions-to-ai-can-reduce-bias/
3. Bianchi, F., Kalluri, P., Durmus, E., Ladhak, F., Cheng, M., Nozza, D., Hashimoto T., Jurafsky D., Zou J. ve
4. Birhane, A., ve Prabhu, V. U. (2021, January). Large image datasets: A pyrrhic win for computer vision?. In 2021 IEEE Winter Conference on Applications of Computer Vision (WACV) (pp. 1536-1546). IEEE. https://doi.org/10.48550/arXiv.2006.16923
5. Bryman, A., ve Bell, E. (2011). Business research methods. Oxford University Press.
6. Caliskan, A. (2023, June). Easily accessible text-to-image generation amplifies demographic stereotypes at large scale. In Proceedings of the 2023 ACM Conference on Fairness, Accountability, and Transparency (pp. 1493-1504).
7. Cao, Y., Zhou, L., Lee, S., Cabello, L., Chen, M., ve Hershcovich, D. (2023). Assessing cross-cultural alignment between ChatGPT and human societies: An empirical study. arXiv preprint arXiv:2303.17466.
8. Coutts, L. (2024). How AI is reshaping power: An urgent call for global governance. LinkedIn. https://www.linkedin.com/pulse/how-ai-reshaping-power-urgent-call-global-governance-lloyd-coutts-cfwjf

9. Deloitte. (2024, July). AI for inclusive development in Africa – Part I: Governance (Research report). Deloitte. https://www2.deloitte.com/content/dam/Deloitte/us/Documents/public-sector/ai-adoption-in-africa-governance.pdf https://www.deloitte.com/content/dam/Deloitte/fpc/Documents/secteurs/technologies-medias-et-telecommunications/deloitte_ai-adoption-africa-2024.pdf
10. Feng, L. (2025). Investigating the effects of artificial intelligence-assisted language learning strategies on cognitive load and learning outcomes: A comparative study. Journal of Educational Computing Research, 62(8), 1961-1994. https://doi.org/10.1177/07356331241268349
11. Feng, X., ve Murakami, M. (2025). Comparative analysis of differentiated approaches to utilizing AI for subverting stereotypes. Journal of Advances in Information Technology, 16(3) 10.12720/jait.16.3.357-371
12. Konda, (2022), Türkiye 100 kişi olsaydı, KONDA İnteraktif, https://interaktif.konda.com.tr/rapor/turkiye-100-kisi-olsa/1#navislide
13. Krippendorff, K. (2004). Content analysis: An introduction to its methodology. Sage publications.
14. Lan, X., An, J., Guo, Y., Tong, C., Cai, X., ve Zhang, J. (2025). Imagining the Far East: Exploring perceived biases in AI-generated images of East Asian women (arXiv preprint arXiv:2504.04865). arXiv. https://arxiv.org/abs/2504.04865
15. Locke, L. G., ve Hodgdon, G. (2024). Gender bias in visual generative artificial intelligence systems and the socialization of AI. AI & SOCIETY, 1-8. https://doi.org/10.1007/s00146-024-02129-1
16. Marinucci, L., Mazzuca, C., ve Gangemi, A. (2023). Exposing implicit biases and stereotypes in human and artificial intelligence: state of the art and challenges with a focus on gender. AI & SOCIETY, 38(2), 747-761. https://doi.org/ 10.1007/s00146-022-01474-3
17. Mayring, P. (2015). Qualitative content analysis: Theoretical background and procedures. In A. Bikner-Ahsbahs, C. Knipping, ve N. Presmeg (Eds.), Approaches to qualitative research in mathematics education: Examples of methodology and methods (pp. 365–380). Springer.
18. Nadeem A, Abedin B ve Marjanovic O (2020) Gender bias in AI: a review of contributing factors and mitigating strategies. Association for Information Systems, 2020. https://aisel.aisnet.org/
19. Neuendorf, K. A. (2017). The content analysis guidebook. Sage.
20. Olier, J. S., ve Spadavecchia, C. (2022). Stereotypes, disproportions, and power asymmetries in the visual portrayal of migrants in ten countries: an interdisciplinary AI-based approach. Humanities and Social Sciences Communications, 9(1), 1-16. https://doi.org/10.1057/s41599-022-01430-y
21. Özkök, E. (4 Ekim, 2018), Kadınların yüzde kaçı başını örtüyor? Türbanlı kadın oranı artıyor mu?, Hürriyet, https://www.hurriyet.com.tr/yazarlar/ertugrul-ozkok/kadinlarin-yuzde-kaci-basini-ortuyor-turbanli-kadin-orani-artiyor-mu-40976040
22. Taubayev, Z., Sultan, Y., ve Dautova, G. (2025). The role of artificial intelligence in forming stereotypes in intercultural communication. Вестник Евразийского Гуманитарного Института, (1), 185–195.
23. Tao, Y., Viberg, O., Baker, R. S., ve Kizilcec, R. F. (2024). Cultural bias and cultural alignment of large language models. PNAS nexus, 3(9),346. https://doi.org/10.48550/arXiv.2311.14096
24. Tibebu, H. (2024, July). Why Africa must demand a fair share in AI development and governance. Austin, TX: Tech Policy Press. Available at: https://www.techpolicy.press/why-africa-must-demand-a-fair-share-in-ai-development-and-governance/
25. Varghese, J. K., ve Rani, P. (2024). Digital Orientalism in machine vision: A cross-platform analysis of AI-generated representations of Indian culture. Социологическое обозрение, 23(4), 113–139. https://doi.org/10.17323/1728-192X-2024-4-113-139
26. Zhang, Y., ve Wildemuth, B. M. (2009). Qualitative analysis of content. Applications of social research methods to questions in information and library science, 308(319), 1-12.
27. Zheng, A. (2024). Dissecting bias of ChatGPT in college major recommendations. Information Technology and Management, 1-12. https://doi.org/10.1007/s10799-024-00430-5