Can AI-generated images match human-generated images in inspiring travel? A visual feature analysis of audience perceptions and visit intention

Generative AI tools can now create photorealistic destination images at low cost, yet whether such images motivate travel as effectively as human-generated images remains unclear. Drawing on the heuristic-systematic model, this research examines differences between AI-generated and human-generated tourism images through three studies. Study 1 shows that AI-generated images receive higher ratings of perceived aesthetics but lower ratings of perceived authenticity, narrative transportation, and visit intention than human-generated images. Study 2 identifies 19 computable visual indicators across three dimensions and finds that AI-generated images exhibit stronger visual schema characteristics, whereas human-generated images show richer visual structure. Study 3 uses XGBoost and SHAP analysis to identify the visual features associated with different psychological responses. These findings connect computational visual analysis with tourism consumer research by identifying differences in audience responses to AI-generated and human-generated images and the visual features associated with those responses.