The Impact of AI on English Pronunciation: A Comparative Study of AI Tools and Traditional Methods

Abstract

The current study explores the effectiveness of AI-driven pronunciation tools in improving the phonetic accuracy and speaking proficiency of English learners. With a growing focus on Artificial Intelligence (AI) in language education, particularly for second language acquisition (SLA), AI-based tools like Bixby and ChatGPT offer the potential for real-time, personalized feedback, addressing critical aspects of pronunciation such as consonant and vowel articulation, stress patterns, and intonation. The research employs a mixed-methods approach, combining quantitative and qualitative data collection methods. A group of English learners from the Department of English and Applied Linguistics at the University of Lakki Marwat, representing different proficiency levels (beginner, intermediate, and advanced), used an AI-powered speech recognition tool that provides immediate, personalized corrections. The findings reveal that the AI tool significantly improved learners' pronunciation accuracy compared with traditional teaching methods, with the most significant gains observed among beginner and intermediate learners. Participants reported higher engagement and motivation from real-time feedback, which contrasts with the slower, less personalized feedback typically offered in traditional classroom settings. However, the study also identified limitations in the AI system's ability to address accent variation, suggesting the need for further refinement in AI algorithms to accommodate a broader range of linguistic backgrounds. Overall, the study concludes that AI-driven pronunciation tools can play a transformative role in SLA by providing efficient, engaging, and individualized pronunciation training. The research also highlights the need for continued development to ensure these tools can be universally effective across different accents and dialects.

Keywords: AI-driven pronunciation tools, SLA, pronunciation accuracy, speech recognition, real-time feedback, personalized learning, English learners, AI, ChatGPT, Bixby

https://doi.org/10.5281/zenodo.18635024