Leveraging the affordances of artificial intelligence chatbots to create differentiated instructions and cater to learners' differences in an English as a foreign language class

Given the rapid evolution in AI, a growing body of literature has investigated the potential benefits it can bring for education in general and language learning in particular. Nonetheless, little research has looked into how to leverage the affordances of AI chatbots to create differentiated instruction (DI) and cater to EFL learners' differences. Thus, we conducted this action research to shed light on 1/ learners’ perceptions about the affordances of AI chatbots as a tool for differentiated instruction, 2/ perceived challenges of integrating AI chatbots in EFL classrooms for differentiated learning, and 3/ stakeholders’ suggestions to leverage AI chatbots to implement differentiated instruction effectively. This study follows a mixed-method approach, with three main data collection instruments, namely students’ survey, students’ follow-up interview, and teachers' reflective journals. Thematic analysis was used to analyze data from the interviews and reflective journals, while quantitative analysis was used for the surveys. Data analysis reveals students’ and teachers’ generally positive perceptions about a variety of affordances provided by AI chatbots, such as quick, personalized feedback, multimodal input. However, different students perceive some of the affordances differently, either as opportunities or difficulties for learning, depending on their learning styles, and most importantly, AI readiness level. Some specific challenges are unearthed, together with corresponding solutions.