The rapid advancement of generative artificial intelligence (GAI) is transforming tourism by enabling the creation of high-quality visual content for destination marketing. While text-based GAI applications are well-studied, the potential of visual-based GAI remains underexplored. This research examines how Destination Marketing Organizations (DMOs) can leverage AI-generated images to enhance destination appeal, using a multi-method approach that combines experiments, deep learning, and big data analysis. Grounded in information processing fluency theory and the destination imagery model, the study reveals that AI-generated visuals are particularly effective for promoting unfamiliar destinations compared to familiar ones. Furthermore, the findings uncover an inverted U-shaped relationship between the consistency of tourists' organic destination image and the perceived attractiveness of AI-generated visuals. These insights provide DMOs with a data-driven framework to optimize the use of AI-generated imagery, enhance marketing effectiveness, and craft compelling visuals that inspire travel interest.