Exploring the Use of Character. AI in Enhancing ESL Learners’ Affective Dimensions

Abstract

ESL learners face challenges in their language learning journey, and such challenges are often caused by communication anxiety, low self-confidence and high-stress conditions. As a result, digital natives like Generation Z ESL learners are increasingly adopting generative AI tools like Character.AI to practise communication in the third spaces. This qualitative study was guided by Krashen's Affective Filter Hypothesis and it explores ESL learners' perceptions of Character.AI as a tool to develop English-speaking skills. Ten non-native ESL speakers were interviewed, and the results were thematically analysed. While participants aimed to develop their speaking fluency, the main mode of interaction via the text-based interface of the platform reveals a pedagogical paradox. Character.AI. Besides lowering learners' affective filters, AI text-based chatbots specifically helped to reduce anxiety among learners, increase their motivation, increase WTC and improve their vocabulary and sentence structure. The findings suggest that Character.AI is a valuable text-based chatbot and a useful preparatory stepping stone for practice and to lower affective filters.