Humans or Generative AI? Influence of recommendation agents on tourists’ decision-making

This study investigates whether Generative AI provides more compelling travel recommendations than humans by exploring the interaction between recommendation agents (Generative AI vs. humans) and temporal distance (near vs. distant). We conducted three scenario-based between-subjects experiments (N = 708, recruited through Credamo, an MTurk-like platform). Participants were randomly assigned to rate their responses to Generative AI- or human-generated recommendations for three scenarios (destination, hotel, and restaurant), with temporal distance adjusted for by scenario description (e.g., “travel next week” vs. “travel next year”). Our results reveal that travel-related temporal distance affects tourists’ preferences for recommendations from humans or Generative AI. Specifically, when tourists plan a near-future (vs. distant-future) journey, human recommendations (vs. Generative AI recommendations) are more persuasive. Further, this interaction effect on tourists’ decision-making is mediated by perceived message credibility and is weakened when recommendation agents make an error. Our findings offer valuable insights into enhancing tourist experiences and promoting tourism products.