Social chatbots should self-correct errors to maintain user trust, study finds

A new study published on arXiv explores how social chatbots can maintain user trust after making errors. Researchers found that when chatbots correct their own mistakes, they retain higher credibility and perceived expertise compared to when an external source corrects them. The study also indicated that a strong social connection between the user and the chatbot amplifies the effectiveness of self-corrections, suggesting that investing in this connection is a functional mechanism for maintaining long-term credibility. AI

IMPACT Suggests design principles for chatbots to build and maintain user trust through effective error correction.

RANK_REASON This is a research paper published on arXiv detailing findings from an experiment on chatbot behavior. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• alphaXiv
• arXiv
• CatalyzeX Code Finder for Papers
• Connected Papers
• CORE Recommender
• DagsHub
• expert chatbot
• Gotit.pub
• Hugging Face
• human–computer interaction
• Influence Flower
• Litmaps
• ScienceCast
• scite Smart Citations
• social chatbots
• webpage retraction

• paper
• safety
• product

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →