This thesis investigated how customers perceive different AI chatbot disclosure strategies after experiencing negative service events, and the impact on customer forgiveness and behavioural outcomes. The empirical research provides suggestions and guidelines for service providers to optimize chatbot integration while ensuring satisfactory service recovery outcomes. Findings show varied customer responses between disclosed and not disclosed chatbots based on specific service failure scenarios. For minor service failures, disclosed chatbots lead to reduced forgiveness but lower intention to spread negative words, whereas there are no differences in customers’ perceptions when service failures are severe.