Generative AI and Human Advice: User Perceptions, Intentions, and Behavior Across Pre-visit and On-site Travel Stages
通过三个实验，研究了用户在前访问和现场旅行阶段对生成式人工智能与人类建议的偏好差异，发现用户在前访问阶段更偏好人类建议，而现场阶段差异不大。
This research explores users’ preferences for human versus Generative Artificial Intelligence at specific stages of the information search process. Through three experiments, we investigate how the source of recommendation – Generative Artificial Intelligence versus Human—shapes user perceptions, intentions, and behaviors. Study 1 shows users favor human providers in terms of perceived information quality, usage intention, and advocacy intentions. Study 2 demonstrates that this preference exists only in pre-visit stages, with no differences for on-site information search (except in willingness to follow recommendations). Study 3 shows that participants are more likely to post positive reviews about human providers pre-visit, with no difference on-site. While previous research has conceptually explored ChatGPT’s role in travel decision-making, our empirical evidence shows that acceptance and advocacy toward Generative Artificial Intelligence technologies shift according to the travel stage, providing valuable insights for tourism information providers.