Dark side of generative AI in tourism: a stressor-strain-outcome perspective; using a mixed-methods approach

Generative Artificial Intelligence (GAI) has revolutionised the creation and dissemination of tourism-related information. To study the challenges experienced by tourists and its influence on their subsequent behaviors, a mixed-method-technique has been used. In phase one, qualitative analysis identifies key challenges (stressors). In contrast, phase two employs a quantitative study to investigate how identified stressors influence tourists’ opinions and behavioral responses. For a qualitative analysis, 44 respondents were interviewed, and data analyzed using Nvivo 11.0 software. Additionally, in phase two, 445 tourists participated in a survey conducted through both online and offline platforms and analysed with structural equation modelling. Findings indicate identified stressors (AI hallucinations, poor information quality, and a lack of empathy) directly impact distrust. Furthermore, all three stressors indirectly contribute to the development of negative e-WOM. Findings also highlight the noteworthy moderating influence of the cognitive image of destination on distrust and negative e-WOM. The necessity for accuracy and transparency is highlighted by the fact that inaccurate recommendations can undermine tourists’ trust and happiness. Focusing on strategies to reduce negative consequences like distrust and electronic word-of-mouth, research offers novel insights into the stressors tourists experience while using GAI for tourism and how it impacts their attitudes and behavior.