Correct Yourself, Keep My Trust: How Self-Correction and Social Connection Shape Credibility in Social Chatbots
arXiv:2606.19286v1 Announce Type: cross Abstract: When social chatbots make mistakes, and they do, how they recover determines whether users trust them again. Social chatbots are increasingly integrated into everyday life, yet they remain prone to generating convincing but inaccurate information. The social connection they build with users makes such errors particularly consequential. We conducted a between-subjects experiment (N=120) comparing three error correction strategies: a webpage retraction, self-correction by the same social chatbot, and correction by an expert chatbot. Our results r
As social chatbots become more integrated into daily life, understanding user trust and effective error correction mechanisms is critical for their continued adoption and societal impact.
The credibility of AI systems, particularly those that build social connections, directly influences user behavior and the responsible deployment of advanced AI applications.
This research provides insights into designing more resilient and trustworthy social AI, moving beyond mere accuracy to incorporate effective recovery strategies post-error.
- · AI developers focused on explainable and trustworthy AI
- · Companies deploying social chatbots
- · Users of social AI tools
- · AI developers ignoring social dynamics and error recovery
- · Chatbots with poor error correction mechanisms
Improved user trust and retention for social chatbots implementing effective self-correction and transparency.
Increased adoption of AI agents in sensitive domains where trust and reliability are paramount.
Potential for new regulatory frameworks focusing on AI transparency, error handling, and accountability in user interactions.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.AI