Should I Use ChatGPT Travel Insurance Recommendations? A Dual-Process Theory Perspective

Abstract

This study employed dual-process theory to examine the behavioral intention to follow ChatGPT's travel insurance recommendations. It investigated the moderating role of insurance literacy on the relationship between the attractiveness of ChatGPT's travel insurance recommendations and the behavioral intention to adopt these recommendations. In total, 378 usable questionnaires were collected through the Prolific platform in the United States. The findings showed that while perceived trust, accuracy, and customization were responsible for the attractiveness of ChatGPT's travel insurance recommendations, perceived convenience and perceived security had no significant effect. In addition, it was found that perceptions of both “coolness” and “creepiness” significantly affected the attractiveness of ChatGPT's recommendations. At the same time, the attractiveness of recommendations influenced the behavioral intention to adopt them. In contrast, insurance literacy did not moderate the relationship between recommendations' attractiveness and the behavioral intention to adopt such recommendations. This study is unique in that it is the first study to employ Dual-process theory in this context. The findings enhance our understanding of which element (cognitive vs. affective) exerts the most substantial influence on the attractiveness of travel insurance based on ChatGPT's recommendations, providing valuable insights into the design and marketing of AI-based travel insurance services.