Promoting Active User Co-Repair in Interaction Failure: User Responsibility Reminder as a GenAI Feedback
Type: article
Abstract: Generative AI chatbots are widely used for complex, open-ended tasks, where interaction failures with causes that are difficult for the chatbot to identify and repair independently are common. Such diagnostic uncertainty requires users and chatbots to collaborate in diagnosing and repairing failures. Drawing on the joint-action view of communication, we propose the User Responsibility Reminder, a form of repair feedback that highlights users’ potential responsibility for failures and thus encourages their active participation in co-repair. We distinguish between brief and detailed URRs and examine their effects in a 3 × 2 between-subjects experiment. Both URRs increased users’ continuance intention by increasing self-attribution and, subsequently, perceived chatbot competence. Brief URR also led to a negative indirect effect by increasing perceived face threat and thereby reducing perceived chatbot competence. User motivation to process the repair feedback further shaped these effects. Our findings provide valuable insights into GenAI failure-recovery designs under diagnostic uncertainty.
Language: English
Topic: AI in Service Interactions
Subfield: Marketing
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7222907833