Artificial Intelligence Tools as Catalysts of Improved Spoken English: A Systematic Review of the Current Applications and Challenges

A goal of English as a Second Language (ESL) learners is to be fluent in spoken English, yet challenges such as pronunciation difficulties, limited fluency, and low confidence persist. With the rise of artificial intelligence (AI), new tools have emerged to support oral language development, and the pedagogical value and limitations of these tools require systematic evaluation. This study conducted a systematic review of empirical research published between 2021 and 2025, was guided by the PRISMA framework and drew on studies in the ERIC and Sage databases. In total 11 studies were analyzed, covering AI applications such as natural language processing-based chatbots (e.g., ChatGPT), AI-powered presentation platforms (e.g., PitchVantage), speech recognition systems (e.g., Speechling, E-platforms), and assessment tools (e.g., Duolingo English Test). Findings show that these tools provide personalized, real-time feedback that enhances pronunciation, fluency, learner autonomy, and engagement. Nevertheless, persistent challenges include the accuracy and precision of feedback, learner dependency on technology, feedback quality and clarity, lack of contextual awareness, technical barriers, and access and inclusive issues, alongside ethical concerns over data privacy. The review concludes that AI tools complement but cannot replace human mediation, and recommends inclusive, context-aware, and ethically governed AI solutions that are integrated with teacher guidance to maximize their effectiveness in developing ESL speaking.