The purpose of this study was to analyse the effectiveness of AI on instruction of English Listening and Speaking among learners with autism in Nakuru East Sub-county and to evaluate the challenges of using AI on instruction of English Listening and Speaking for learners with autism in Nakuru East Sub-county. The study was guided by the Usage-Based Theory of Language Acquisition, proposed by Michael Tomasello. It employed the quasi experimental mixed-methods research design. The target population of the study was 200 learners with autism in 6 special needs schools in Nakuru, from which a sample of 133 learners was selected through simple random sampling. Additionally, 6 teachers were selected for interviews. Data was collected using a language instruction test for learners and an interview guide for the teachers. Quantitative data was analyzed using descriptive and inferential statistics, while qualitative data from semi-structured interviews with teachers was thematically analyzed. Learners’ post-intervention language performance was not strongly linked to their initial ability levels (B = 0.007, p = .923). Learners exposed to AI tools demonstrated stronger gains compared to those taught using traditional methods. The experimental group consistently outperformed the control group even after accounting for initial differences. Learners in the AI group had more frequent transitions from incorrect to correct responses. The study concludes that AI-supported language instruction is more effective than traditional teaching methods. Teacher training institutions and the Teachers Service Commission should enhance CPD initiatives that are oriented towards integrating AI in the teaching of languages.
Keywords: Artificial Intelligence, Effectiveness, Nakuru east sub-county, language instruction, learners with autism
