This study examines how guests interact with artificial intelligence (AI) technologies in tourism and hospitality settings using actor-network theory (ANT), addressing critical gaps in understanding AI adoption patterns, variations in guest satisfaction and the formation of stable human-technology networks across different traveler segments and service touchpoints.
The research analyzed 20,000 TripAdvisor guest reviews from January 2023 to December 2024 using a mixed-methods approach. Qualitative thematic analysis via NVivo 14 identified 13 distinct technology themes, while python-based natural language processing employed VADER sentiment analysis.
AI contactless payments (88% adoption) and digital keys (82% adoption) demonstrated stable actor-networks, while AI chatbots showed critical instability with 48% negative sentiment and declining trust (3.8/10). Hybrid human–AI service channels achieved the highest satisfaction (8.4/10) compared to fully automated systems. Traveler preferences varied considerably, from 92% AI preference among tech enthusiasts to 18% among seniors. Critical unmet needs emerged, including luggage tracking (9.1 pain level) and discovery of authentic experiences (8.5 pain level), representing opportunities for AI intervention.
This study offers a novel empirical application of ANT, supported by large-scale review analytics, operationalizing its core constructs of translation and enrollment. It provides empirical evidence of task-specific AI agency in tourism and hospitality, showing guests grant agency to invisible AI enhancements while resisting conversational AI replacements for human service.
