Investigating the Persuasive Potential of Communicating With Generative Artificial Intelligence Chatbots for Mental Health: The Roles of Perceived Message Contingency, Cognitive Elaboration, and Issue Involvement

With the ever-increasing adoption of generative artificial intelligence (GenAI) chatbots in diverse communication roles, there is a burgeoning need to address how people process and respond to messages from these technologies. This study explores the theoretical mechanisms through which the use of GenAI chatbots influences persuasive outcomes related to mental health, examining perceived message contingency and cognitive elaboration as mediators, and issue involvement as a moderator between the two mediators. Through an experiment using ChatGPT, this study found that, compared to non-conversational online resources, the use of a GenAI chatbot led to greater cognitive elaboration via an increased sense of message contingency, ultimately leading to greater mental health self-efficacy. The influence of perceived message contingency on cognitive elaboration was more pronounced among participants with higher issue involvement.