Artificial Intelligence (AI) in Enhancing Spoken English Proficiency: A Systematic Literature Review

International Journal of English Linguistics, Literature, and Education (IJELLE) Vol. 6, No. 2, December 2024, pp. 156-169 ISSN 2686-0120 (print), 2686-5106 (online) 156 http://journal.univetbantara.ac.id/index.php/ijelle/index Artificial Intelligence (AI) in Enhancing Spoken English Proficiency: A Systematic Literature Review Ainol Mardhiah a,1,*, Diana Purwati b,2, Lathifatuddini c,3, Hayatul Muna d,4, Helmiyadi a,5 a English Education Department, Universitas Bumi Persada, Lhokseumawe 24351, Indonesia Education Department, STKIP Yapis Dompu,, NTB 84216, Indonesia c English Education Department, STIT Darussalam, Lhokseumawe 24351, Indonesia d English Education Department, IAIN Lhokseumawe 24351, Indonesia 1 *; 2 ; 3 ; 4 ; 5 * Corresponding Author b English Received 23 November 2024; accepted 30 December 2024; published 31 December 2024 ABSTRACT This systematic review explores the use of Artificial Intelligence (AI) tools in enhancing spoken English proficiency among EFL learners, focusing on intelligent tutoring systems, speech recognition applications, and adaptive feedback. Using the PRISMA framework, 38 relevant articles (2020–2024) were identified through Google Scholar searches, with seven additional articles found manually. The review included full-text studies involving higher education participants and AI-based interventions for spoken English, excluding books, theses, and unrelated research. The data was then analyzed using thematic analysis highlighting the benefits and challenges of AI tools. Intelligent Tutoring Systems (ITS), such as chatbots and virtual tutors, improved learners’ fluency, pronunciation, and confidence through personalized experiences. Platforms like Duolingo, Alexa, and Replika enhanced engagement but struggled with emotional factors like speaking anxiety and often provided generalized feedback. AI-driven speech recognition tools reduced anxiety, improved pronunciation, and boosted motivation but faced issues with diverse accents and conversational nuances. AI feedback mechanisms delivered real-time, personalized feedback to enhance proficiency, though accuracy and adaptability to diverse learners remain challenges. AI tools show strong potential for enhancing spoken English learning through personalized and interactive methods. However, addressing emotional engagement, learner diversity, and accessibility is crucial for broader adoption in higher education settings. KEYWORDS Artificial Intelligence (AI) EFL Higher Education Spoken English Proficiency This is an openaccess article under the CC–BY-SA license 1. Introduction In recent era, English has grown to play a crucial role in achieving success across various domains, such as education, employment, and global communication (Lan et al, 2020). Among the fundamental competences in language learning, spoken English remains an important area of focus, particularly for students in English as a Foreign Language (EFL) contexts (Grabe & Stoller, 2002). However, achieving spoken English proficiency presents numerous challenges for learners, especially in contexts where exposure to English outside the classroom is limited. Factors such as insufficient opportunities for authentic communication, anxiety about making mistakes, and irregularities in English pronunciation and spelling often hinder students' progress (Cumming et all, 2018; Tokoz-Goktepe, 2014; Ulicheva et al, 2018). Artificial Intelligence (AI) has increasingly become a transformative tool in language education, introducing innovative solutions to address challenges in English teaching and learning (Baranwal, 2022; Kannan & Munday, 2018). AI entails a range of technologies, such as intelligent tutoring systems, speech recognition software, as well as adaptive feedback mechanisms, which simulate human cognitive functions and provide personalized learning experiences (Sindermann, 2021; Xiaohong & Yanzheng, 2021). These AI tools are being increasingly utilized in English language education to help the improvement of spoken English skills. Research has shown that AI doi :10.32585/ijelle.v6i2.5980 157 International Journal of English Linguistics, Literature, and Education (IJELLE) Vol. 6, No. 2, December 2024, pp. 156-169 ISSN 2686-0120 (print), 2686-5106 (online) can significantly enhance different aspects of language learning, including improving pronunciation accuracy and boosting learner engagement through interactive and adaptive learning platforms. (Kim, 2019; Noviyanti, 2020; Pokrivčáková, 2019). In contexts where exposure to English outside the classroom is limited, AI tools provide crucial opportunities for extended practice and skill development, offering a practical solution to overcome challenges in learning English. Artificial Intelligence (AI) has significantly impacted language learning and teaching. AI technologies that have emerged in this field include intelligent tutoring systems, speech recognition software, and adaptive feedback mechanisms. These tools are designed to simulate aspects of human cognitive processes, enabling personalized and efficient learning experiences. Intelligent Tutoring Systems (ITS) are AI-powered platforms which are designed to offer instruction and feedback based on individual learner progress, thereby enhancing language learning through targeted activities and resources (Crompton et al, 2024). Intelligent tutoring systems in EFL contexts offer students a platform for extensive, personalized practice based on their proficiency level. These systems are especially advantageous in contexts with limited access to proficient English speakers, as they emulate interactions with native speakers through AI-driven dialogues and scenarios (Kang, Jeon & Lee, 2024). Such immersive experiences play a critical role in fostering the communicative competence essential for achieving spoken English proficiency. Intelligent Tutoring Systems (ITS) in EFL settings include platforms such as ELSA Speak, Smalltalk, D-ID Studio, Duolingo, Replika, and HelloTalk. These tools leverage AI to create interactive conversations, deliver pronunciation correction, and tailor learning to individual needs. ELSA Speak, for example, focuses on enhancing pronunciation with instant feedback, while Duolingo gamifies lessons to gradually improve language skills. Replika and HelloTalk enable users to engage in text or voice chats with AI or native speakers, providing authentic practice opportunities. These applications play a crucial role in offering personalized and engaging language practice, supporting learners in developing their English proficiency. Humans convey information using words, which are made up of phonemes—the fundamental sounds of language. The process of generating artificial speech begins by identifying the intended message and translating it into language (Sutrisno, 2018). Motor neuron signals are then sent to the vocal cords, prompting them to vibrate and produce sound waves, which are carried to t (...truncated)