Do Chatbot Attributes Influence Perceived Creepiness and Paradoxical Behavior? The Role of Trust in Social Media Influencers
(A2026-134349)

AUTHORS

Tahir Islam, Leeds Trinity University/AGH University of Krakow; Naeem Akhtar, AGH University of Krakow; Dariusz Cichon, AGH University of Krakow; Rauf Islam, University of Electronic Science and Technology of China; Abdul Ghaffar, University of Karachi; TOMÁŠ KINCL, Faculty of Management, Prague University of Economics and Business

ABSTRACT

This study explores the Chatbot conversational attributes and developed and tested a research framework using the SOR model as theoretical foundation. We examine how the Chatbot attributes namely competence, warmth, task complexity, and usability develop users’ creepiness, result in perceived Chatbot value. We assess the role of perceived Chatbot value in developing paradoxical behaviors—repeat patronage intentions and choice deferral. Social media influencers’ trust moderates the effects of perceived Chatbot value on paradoxical behaviors. Using representative samples from US tourists, data was collected from 316 respondents via Prolific Academic platform. We performed PLS-SEM to analyze data in Smart PLS 4.0. Results indicated that Chatbot conversational attributes significantly develop users’ creepiness. We found that users’ creepiness negatively develops perceived Chatbot value, leading positively repeat patronage intentions and negatively choice deferral. Social media influencers’ trust significantly moderates the relationship between perceived Chatbot value and repeat patronage intentions and insignificant choice deferral. This research theoretically and practically extends the literature on tourism sector, latest AI-enabled applications, consumer psychology, and online booking platforms. We acknowledge a few limitations and offer directions for upcoming studies. Keywords: Psychology, competence, warmth, perceived value, choice deferral, repeat patronage intentions