Enhancing cross-border shopping through AI chatbots: an antecedents–engagement–consequences perspective of brand-consumer interactions

Purpose This study aims to develop and test an Antecedents–Engagement–Consequences model investigating how generative artificial intelligence (AI) chatbots influence cross-border shopping. This study examines how perceived usefulness, ease of use, competence, quality and threat affect consumer engagement and subsequent shopping outcomes by using perceived warmth as a moderator. Design/methodology/approach The study uses a two-wave time-lagged survey design with 276 consumers using the AI Jingdong Jingyan chatbot on Jingdong Worldwide. Data analysis combines partial least squares structural equation modeling and fuzzy-set qualitative comparative analysis (fsQCA) approaches, supplemented with semi-structured interviews for deeper insights into consumer–AI brand interactions. Findings Perceived usefulness, ease of use, competence and quality significantly enhance chatbot engagement, driving international shopping frequency and satisfaction. Perceived warmth moderates these relationships, while perceived threat shows no significant effect. Multiple configurational paths to high shopping frequency emerge through fsQCA analysis. Practical implications The findings guide e-commerce platforms in optimizing AI chatbot design and functionality to enhance brand–consumer interactions. Recommendations include strengthening personalization capabilities and developing warm interaction styles to boost engagement and satisfaction. Originality/value This study pioneers an integrated framework to examine AI chatbot-mediated brand interactions in cross-border e-commerce, extending consumer engagement literature and Technology Acceptance Models. It offers novel insights into how AI chatbots can strengthen brand–consumer relationships across cultural and linguistic barriers.