Double-sided messages improve the acceptance of chatbots

Customers may not wish to use an artificial intelligence (AI) chatbot after knowing its nonhuman identity. Based on the inoculation theory, this study explores how to increase customers' willingness to communicate with AI chatbots by adopting a double-sided message strategy after AI's nonhuman identity is disclosed through one field experiment in a hotel and three online vignette experiments. Results indicate that a double-sided message strategy enhances customers' willingness to interact with AI chatbots via the mediating role of perceived authenticity. Furthermore, the study highlights that customers' attitudes toward interacting with AI chatbots are moderated by customer demand types—complaints and inquiries. This study proposes an effective solution to customers' rejection of AI chatbots.