Jul 2026· International journal of social science and human research· 0 citations
TL;DR
Across the reviewed studies, generative AI was found to enhance language learning through personalized feedback, increased learner autonomy, and greater learning engagement, but concerns regarding academic integrity, AI literacy, ethical issues, and institutional readiness remain significant challenges to its sustainable implementation.
Abstract
The rapid development of generative artificial intelligence (AI) has significantly transformed university English as a Foreign Language (EFL) education, creating new opportunities for language learning, teaching, and assessment. This review synthesizes recent studies published between 2023 and 2025 to provide an overview of current research trends, major findings, research gaps, and pedagogical implications of generative AI in university EFL contexts. A critical narrative review with a systematic literature search was conducted using studies retrieved from Scopus, Web of Science, and Google Scholar. The reviewed literature indicates that research has expanded rapidly following the emergence of ChatGPT, with writing instruction, learner perceptions, and language performance being the most frequently investigated topics. Across the reviewed studies, generative AI was found to enhance language learning through personalized feedback, increased learner autonomy, and greater learning engagement. However, concerns regarding academic integrity, AI literacy, ethical issues, and institutional readiness remain significant challenges to its sustainable implementation. The review also identifies several research gaps, including the limited diversity of AI applications investigated, the dominance of short-term quantitative studies, and the lack of longitudinal and classroom-based research. Overall, this review highlights the importance of integrating generative AI through responsible pedagogical practices and provides directions for future research and AI-enhanced university EFL education.
Artificial intelligence (AI) is increasingly shaping higher education, particularly in the development of academic writing and speaking skills. While AI tools offer immediate feedback and personalized learning opportunities, existing research often focuses on their effectiveness without fully addressing their pedagogical and ethical implications. This creates a need for a more critically informed understanding of how AI influences language learning. This study examines the role of artificial intelligence (AI) in enhancing both academic writing and speaking skills in higher education through a systematic review of recent empirical studies. Drawing on 109 studies published between 2022 and 2025, the review adopts PRISMA guidelines to identify trends in the use of these tools. The findings indicate significant benefits, including increased learner engagement, improved linguistic accuracy, and immediate individualized feedback. These benefits include lexical development, structural coherence, improved pronunciation, and increased learner confidence through iterative practices. However, the review also identifies critical challenges, including risks of overreliance, reduced learner autonomy, and concerns related to linguistic bias. To address these concerns, the study proposes the implementation of the Mediated AI-Pedagogy Cycle, which positions educators as mediating agents between AI affordances and learner development. The study contributes a pedagogically grounded framework for integrating AI into higher education language instruction.
Abdullah Alazemi, Abdullah A. Alenezi, Amer Alsouyan· Education sciences· 0 citations
This paper examines the existing research on generative artificial intelligence (GenAI) in language education and highlights its potential to reshape teaching practices, learner engagement, and pedagogical design. By leveraging co-word analysis and BERTopic modeling on 908 publications from 2023 to the end of 2025, the article traces thematic patterns and conceptual developments in the GenAI field. The co-word analysis identifies key clusters at the intersection of GenAI and English language instruction, including its use in academic writing, translation, assessment, and learner-centered pedagogy. These themes demonstrate the critical role of GenAI in enabling personalized feedback, adaptive learning environments, and enhanced teacher–student interaction. The application of BERTopic modeling adds a semantic layer to this analysis and reveals diverse topics such as AI-assisted writing, teacher engagement with AI tools, emotional and motivational dynamics, and technology acceptance among learners. Results also indicate a progression from broad pedagogical experimentation toward more focused inquiries into learner psychology, ethical use, and professional development. This trend reflects the maturation of GenAI applications in both formal and informal educational contexts. The findings offer implications for language educators, curriculum designers, and policymakers by highlighting the need to integrate GenAI through pedagogically grounded, ethically responsible, and learner-centered approaches. They also suggest that future language education practices should combine AI literacy, transparent assessment policies, and teacher professional development to support effective and equitable GenAI adoption. Ultimately, the article positions GenAI as a transformative force in language education that offers technological innovation and new frameworks for inclusive, responsive, and emotionally attuned instruction.
Abderahman Rejeb, K. Rejeb, Heba F. Zaher et al.· Quality & Quantity· 0 citations
The findings indicate that GenAI enhances personalized learning, learning motivation, academic engagement, self-regulated learning, self-regulated learning, and learning effectiveness while supporting more flexible and learner-centered educational environments.
Dương Khánh Nguyễn, Liem Van Bui, L. Quynh· Tennessee Community Service...· 0 citations
It is proposed that mathematical literacy in the AI era should extend beyond traditional competencies to include the ability to critically evaluate AI-generated outputs, identify algorithmic limitations, and use AI responsibly a construct this review tentatively terms Mathematical AI Literacy.
Andi Mangaraja, D. Hasibuan, Ramadhan Herianto et al.· Mathline : Jurnal Matematika...· 0 citations
Key challenges included algorithm reliability, institutional readiness, data privacy, academic integrity, academic integrity, and equitable implementation, highlighting the need for human-centered, ethically guided AI integration.
Muhammad Imran, N. Almusharraf, Khurram Shehzad· Journal of Interdisciplinary...· 0 citations
An AI-Mediated Learning Culture Framework is proposed that maps the interrelationships between technological affordances, academic practices, and learner agency and offers directions for empirical research and supports higher education institutions in designing AI-responsive learning environments.
Rachmat Satria· IQRO Journal of Islamic Educ...· 0 citations