How AI-Based Recommendations on Short Video Platforms Drive Tourists’ Decisions: A Cross-Cultural Study of Destination Authenticity from Vietnam and France

Abstract

This study examines factors influencing tourist perceptions and intentions toward artificial intelligence (AI)-recommended destinations on short-form video platforms, specifically how algorithmic characteristics and content attributes impact destination perceptions of authenticity, credibility, and advocacy in Vietnam and France by using the Elaboration Likelihood Model (ELM) and the Stimulus-Organism-Response (SOR) framework. Data was collected from 729 tourists using short video platforms (481 from Vietnam, 248 from France) via structured questionnaires. Analysis through Partial Least Squares Structural Equation Modeling (PLS-SEM) revealed notable cross-cultural differences: algorithmic unbiasedness and diagnosticity positively affect authenticity perceptions, particularly among Vietnamese users who prioritize fairness and personalized relevance. In contrast, French users are significantly influenced by content attributes like accuracy and entertainment, which enhance credibility and advocacy intentions. Additionally, perceived destination authenticity mediates the relationship between algorithmic features and user intentions, affecting credibility, advocacy, and visit intentions in both countries. This research enriches existing tourism and AI technology literature by incorporating under-examined dimensions such as algorithmic unbiasedness and self-compatibility in various cultural contexts, providing valuable insights for tourism marketers and policymakers to enhance user engagement and foster authentic travel experiences through tailored AI algorithms and content strategies on emerging social media platforms.