Hao Wu

Proceedings of the 16th International Conference on Society and Information Technologies: ICSIT 2025, pp. 75-77 (2025); https://doi.org/10.54808/ICSIT2025.01.75

Hao Wu
School of Education, University of Saint Joseph, Macau

Cite this paper as:
Wu, H. (2025). Assessing Students' Ability to Identify AI-Generated Images: A Two-Dimensional Study of Realistic and Artistic Content. In N. Callaos, J. Horne, B. Sánchez (Eds.), Proceedings of the 16th International Conference on Society and Information Technologies: ICSIT 2025, pp. 75-77. International Institute of Informatics and Cybernetics. https://doi.org/10.54808/ICSIT2025.01.75

With the rapid development of artificial intelligence (AI), especially generative AI, there is increasing concern regarding the risk of misinformation from AI-generated content. This study designed a two-dimensional instrument to assess students' ability to identify AI-generated images in both realistic and artistic contexts. The sample consisted of a group of university students (n=112). The instrument included 8 images (4 realistic and 4 artistic), and students were required to identify whether the image was AI-generated. The instrument demonstrated acceptable reliability and good construct validity. The results suggested that students' abilities to identify AI-generated images are two independent constructs but are positively correlated. Additionally, the findings indicated that it is particularly challenging for students to identify AI-generated images in the artistic context. The results highlight the importance of providing training to students in this area, laying a foundation for future studies.