Digital transformation in Islamic banking has accelerated the adoption of artificial intelligence-based customer service, including chatbot systems that are expected to provide fast, accessible, and inclusive banking assistance. This study aims to evaluate the user experience of Aisyah Chatbot as an English-speaking virtual assistant in Islamic banking services by focusing on linguistic quality, UX writing, conversational relevance, and language-related communication breakdowns. This study employed a descriptive qualitative approach using simulated English chat transcripts as the primary data source. The data were analyzed through an interactive model consisting of data reduction, data display, and conclusion drawing, supported by UX writing principles, Gricean conversational maxims, and linguistic breakdown analysis. The findings indicate that Aisyah Chatbot performs relatively well in informational and educational service scenarios, particularly when users ask clear and formal questions about Islamic banking products and digital account services. However, weaknesses remain in complaint-handling scenarios, especially when users employ natural English, informal expressions, broken grammar, or context-specific problem statements. These weaknesses appear in the form of irrelevant fallback responses, limited understanding of Islamic banking terminology, insufficient empathy, and failure to provide actionable solutions. The study concludes that language quality is a central determinant of chatbot user experience in Islamic banking because linguistic accuracy, pragmatic relevance, and empathetic response design directly influence service efficiency, user trust, and customer satisfaction. The implication of this study is that Islamic banking chatbots should be optimized through stronger English corpora for sharia banking terminology, improved natural language understanding, more contextual fallback responses, and seamless handoff mechanisms to human customer service when automated responses fail.
Copyrights © 2026