Chatbots and service failure: When does it lead to customer aggression

Artificial intelligence technology is changing the way services are delivered and introducing opportunities for new sources of service failure. The purpose of this paper is to examine how customers might respond (emotion- or problem-focused coping) to service failure of a chatbot when there is an option to interact with a human employee. Using data from 145 participants, we found that in a chatbot service failure context, telling a customer late in the service interaction that a human employee is available to help leads customers to engage in emotion-focused coping, resulting in customer aggression. The positive relationship between late disclosure and emotion-focused coping occurs with those who perceive low customer participation whereby they do not believe they are overly involved in co-producing and co-delivering the service. This research demonstrates how chatbot service failure in a service encounter can produce different effects on customers’ intention to engage in aggression.