An Evaluation of Generative AI Chatbots Based on Usability and Transparency

This thesis aims to examine and compare different generative AI chatbots on mobile platforms in order to provide design recommendations for future chatbot design processes. The evaluation focuses specifically on usability and transparency. While traditional technology adoption frameworks identify usability as a key factor influencing users’ adoption of new technologies, transparency has emerged as an additional determinant, particularly in the context of chatbot systems. To achieve the objectives of this study, five chatbots were selected based on predefined criteria and systematically evaluated across the targeted dimensions using a structured protocol. This protocol was derived from an existing checklist developed in previous research to assess chatbot user experience. The findings indicate that none of the evaluated systems significantly outperformed the others; rather, each demonstrated strengths in specific areas, offering valuable insights for future chatbot design. These findings are discussed in detail, and design recommendations are provided to inform the development of future chatbots.