Visual Divide in Tourism: Multi-Study Analysis of Framework, Differences, and Impacts of GAI Versus Human-Generated Images

Despite the substantial attention given to generative artificial intelligence (GAI), a unified dimensional framework and research on the differences and impact between GAI and human-generated images is still lacking. This study comprises three sub-studies. Study 1 combines a literature review and an expert elicitation approach to construct a visual analysis framework in tourism. Study 2 selects four GAI models to generate 20,000 images, which are then compared with 18,647 human-generated images from Flickr, and finds that visual attributes and quality metric represent the core differences between the two groups. Based on novelty categorization theory and schema congruity theory, study 3 constructs a conceptual model of how visual homogeneity affects tourist travel intentions through perceived creativity and authenticity. Results find that high visual homogeneity significantly reduces perceived creativity and authenticity. This study enriches current theories of tourism image research and provides practical references for tourism practitioners and artificial intelligence developers.