How GenAI travel assistants’ interactions affect user trust and consumer empowerment: the moderating role of perceived manipulative intent
(A2026-133792)

AUTHORS

Luisa Andreu, University of Valencia; Enrique Bign�, Universitat of Valencia; Anna Mattila, School of Hospitality Management, Pennsylvania State University, University Park, Pennsylvania, USA; Iis Tussyadiah, University of Surrey

ABSTRACT

The use of GenAI as travel assistants enables unprecedented levels of personalization, but it can also raise concerns about deceptive practices and manipulative intent. Drawing on the serendipity concept and persuasion knowledge model, this study examines the impact of positive and deceptive GenAI interactions on trust and consumer empowerment, and the moderating role of perceived manipulative intent. Using a simulated Travel assistant ChatGPT to perform a real task, we conducted a between-subjects experiment (nP= 241 and nD= 243) with a single-factor (type of GenAI: positive vs deceptive). Using one-way ANOVA and PROCESS, the results show that negative GenAI generates lower empowerment. Perceived trust in GenAI interactions mediates the relationship between the GenAI type and consumer empowerment. Perceived manipulative intent moderates the effect of GenAI interactions on perceived trust in the ChatGPT interaction. These findings provide valuable insights for both academia and practitioners, as well as social implications regarding the use of dark patterns.