Developing learners’ communicative competence in English as an international language using non-native English variations with AI speech synthesis technology

[EN] This study developed an English as an International Language (EIL) learning program that integrates non-native English variations generated through AI-driven text-to-speech (TTS) technology and examined its educational effectiveness through a pilot intervention. The program was conducted with 15 Japanese university students over five sessions. To assess its impact, pre- and post-questionnaires were administered, including a listening test using TTS voices and the EIL Perception Scale, along with open-ended reflective responses. Although no statistically significant improvement was found in listening test scores, a positive trend was observed in the Strategies for Multilingual/Multicultural Communication subscale. Qualitative responses further revealed that the students developed more open and appreciative attitudes toward diverse English accents. These results suggest that exposure to TTS-synthesized non-native English variations can foster awareness of linguistic diversity and encourage learners to reframe native-speaker norms. However, limitations such as the short duration of the intervention, the small sample size, and reliance on self-report measures highlight the need for caution in interpretation. Future research should involve longer-term interventions, larger and more diverse learner populations, and multimodal assessment tools to more comprehensively evaluate the pedagogical potential of AI TTS-enhanced EIL education.