Skip to main navigation Skip to search Skip to main content

ChatGPT as a Digital Dining Companion: Examining AI Perceptions and Consumer Loyalty in Restaurant Recommendations

  • Minseong Kim
  • , Sae Mi Lee*
  • *Corresponding author for this work
  • Louisiana State University in Shreveport

Research output: Contribution to journalJournal articlepeer-review

Abstract

Growing artificial intelligence (AI) use in hospitality reshapes consumer decision-making, particularly through AI-driven recommendation systems like ChatGPT. Previous research has largely emphasized the functional aspects of AI recommendations; therefore, this study investigates the psychological and emotional mechanisms that drive consumer loyalty toward ChatGPT for restaurant recommendations. Grounded in the Stimulus–Organism–Response framework, this study examines perceived warmth, communication speed, and perceived competence as key stimuli influencing consumer experiences. Trust, attitudes toward ChatGPT, parasocial interaction, and psychological ownership mediate consumer loyalty. To ensure a qualified sample, data were collected from 528 participants through a marketing research firm. The findings highlight the significance of both cognitive and emotional dimensions in AI-human interactions, demonstrating that trust and psychological engagement enhance consumer loyalty to AI-driven recommendations. This study advances AI perception frameworks and provides practical insights for marketers and AI developers seeking to optimize AI-driven recommendation systems to foster deeper consumer relationships.

Original languageEnglish
JournalInternational Journal of Human-Computer Interaction
DOIs
StateAccepted/In press - 2026

Keywords

  • AI-driven recommendation systems
  • ChatGPT
  • consumer loyalty
  • parasocial interaction
  • psychological ownership

Fingerprint

Dive into the research topics of 'ChatGPT as a Digital Dining Companion: Examining AI Perceptions and Consumer Loyalty in Restaurant Recommendations'. Together they form a unique fingerprint.

Cite this