Generative artificial intelligence (GenAI) is increasingly regarded as a transformative technology (Dwivedi et al., 2023; Kilian et al., 2023), prompting growing academic attention towards understanding how teachers integrate this technology into teaching and learning practices (Holmes et al., 2019). Research indicates that the successful integration of GenAI in school contexts depends not only on the affordances of the technology, but on how teachers perceive, adapt, and mediate its use in classroom contexts (Prestridge, 2024). The aim of this paper is to explore how teachers integrate GenAI into their teaching practices in mainstream classrooms to support students with English as an Additional Language (EAL) and to consider the evidence regarding its impact on these students’ learning outcomes.

Globally, education systems are increasingly implementing GenAI through policy, funding, and digital infrastructure reforms (OECD, 2023). In Australia, for example, one in five schools reports increased use of GenAI-related technologies (Williams, 2025). Such rapid growth reflects education systems’ recognition of GenAI’s potential to strengthen learning outcomes, while reinforcing its position as a transformative force within the education sector (Kilian et al., 2023). GenAI tools capable of generating text, images, explanations, and personalised feedback are becoming increasingly integrated into classroom teaching, particularly as schools explore their capacity to support differentiation, scaffolded learning, and student engagement (Chiu et al., 2023), yet their impact remains uneven across learner groups.

Within educational research, GenAI is not only regarded as a technological innovation, but also a pedagogical resource with particular relevance for EAL students (Kilian et al., 2023).  Studies indicate that GenAI’s supports scaffolding (Hsiao & Chang, 2023; Woo et al., 2023), differentiation (Huang et al., 2023; Lin et al., 2025; Tai & Chen, 2024), and targeted language support (Nazari et al., 2021; Xia et al., 2022), enabling improved access to curriculum content and supporting language proficiency. However, the extent to which EAL students benefit from GenAI depends on teachers’ pedagogical knowledge, instructional approaches, digital competence, and confidence in using GenAI responsibly and accurately (Chiu et al., 2023; Fawns, 2022; Gibbons, 2014; Mancusi, 2024; Prestridge, 2024). Without these capabilities, GenAI risks reinforcing educational inequities, particularly for EAL students who require explicit, linguistically informed scaffolding.

In this paper, I draw on the analysis of 11 empirical studies to explore how teachers implement GenAI in mainstream classrooms that include EAL students, and how these practices shape learning and language outcomes.  The systematic review of literature addressed two key questions: (1) How do secondary mainstream teachers integrate GenAI into their teaching practices to support EAL students?” and (2) “What empirical evidence evidence exists regarding GenAI’s impact on student learning?”.Across the literature, student outcomes were linked to teachers’ pedagogical decision-making, professional confidence, and ability to integrate GenAI within purposeful classroom practices (Chiu et al., 2023; Tai & Chen, 2024; Khan et al., 2024; Woo et al., 2024). Teachers, therefore, play a central role in shaping students’ engagement with GenAI, ensuring its use is safe, culturally responsive, and aligned with learning goals.

A systematic review methodology was selected to ensure the rigour and reliability of the study’s conclusions, and to minimise bias throughout the review process (Gough et al., 2017). Systematic reviews are appropriate for synthesising evidence in evolving fields such as generative AI (GenAI). To further promote transparency and accuracy, the review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) framework (Page et al., 2021). The literature search was conducted in April 2025 using three electronic databases: ProQuest Education, Scopus, and EBSCOhost (ERIC). All databases were selected because they provided a range of peer-reviewed literature across relevant areas in education, technology integration, and language learning. This ensured comprehensive coverage of studies related to GenAI, EAL instruction, and mainstream secondary teaching contexts. Given the recent integration of GenAI in education, the search was restricted to studies published between 2015 and 2025. Only peer-reviewed journal articles published in English were included, ensuring academic quality and methodological rigour. Search terms were developed and included combinations of keywords related to artificial intelligence, teacher, learner, secondary school, and EAL. Following duplicate removal, titles and abstracts were screened for relevance by the first author in the first phase of screening. In the second screening phase, two additional authors independently reviewed 93 titles and abstracts against the inclusion criteria. This process resulted in 21 potentially eligible studies, comprising: ProQuest (n = 9), Scopus (n = 5), and EBSCOhost (ERIC) (n = 7). Studies excluded at this stage did not sufficiently address classroom-based GenAI use or its relevance to EAL student learning. Minor discrepancies between reviewers were resolved through discussion, after which the lead author conducted a final consistency check. The final step involved full-text screening of the 21 articles, all of which were read in their entirety to determine whether they presented empirical findings and explicitly addressed the research questions. Several studies were excluded at the full-text stage for clearly defined reasons, mostly because they focused on policy, theoretical, ethical, or organisational perspectives rather than teachers’ classroom practices or EAL student learning outcomes (e.g., Grieve et al., 2024; Haim et al., 2022; Pourdana & Tavassoli, 2022). This process resulted in a final sample of 11 empirical studies. These studies were analysed using narrative synthesis, which involved systematically comparing and interpreting findings across studies to identify patterns.

This review examined how secondary mainstream teachers use GenAI to support EAL students and synthesised the empirical evidence regarding how it influences EAL students’ learning. The findings indicate that when classroom teachers purposefully integrate GenAI into their teaching practices, it has the potential to improve EAL students’ language acquisition, strengthen literacy skills, and promote more flexible and adaptive learning. However, many studies also emphasised the existing challenges, including teachers’ limited digital literacy, lack of confidence, and reliance on GenAI for low-level tasks. Although several studies indicate that GenAI can support EAL students’ learning through differentiation, scaffolding, and adaptive language learning, few have examined specifically how these strategies are most effective for EAL learners. Given that EAL students require explicit linguistic and cognitive support, future research should investigate how GenAI can be aligned with established EAL pedagogical frameworks to improve learning outcomes. Much of the research focused on basic GenAI applications, such as translation and grammar correction, rather than deeper pedagogical uses. As a result, there remains limited empirical understanding of how GenAI may support language development, differentiation, or subject-specific learning. This gap highlights the need for research that moves beyond AI-based applications to examine teachers’ instructional decision-making and its role in improve technological, linguistic, and academic outcomes for EAL students. The findings of this review suggest that GenAI can be meaningfully and effectively integrated into mainstream classrooms when teachers are supported through targeted professional learning, clearly defined pedagogical frameworks, and opportunities for ongoing collaborative reflection. From a practical perspective, the development of teachers’ GenAI pedagogical knowledge, together with the ethical, critical, and responsible use of GenAI, should be embedded into teachers’ ongoing professional practice. Such an approach ensures that GenAI complements rather than replaces effective teaching practices and supports the linguistic, cognitive, and learning needs of EAL students.

Chiu, T. K. F., Moorhouse, B. L., Chai, C. S., & Ismailov, M. (2023). Teacher support and motivation with AI chatbots; Dwivedi, Y. K., Kshetri, N., Hughes, L., Slade, E. L., Jeyaraj, A., Kar, A. K., et al. (2023). ChatGPT and educational implications; Fawns, T. (2022). Entangled pedagogy beyond technology; Gibbons, P. (2014). Scaffolding language learning; Gough, D. (2017). Introduction to systematic reviews; Grieve, A., Rouhshad, A., Petraki, E., Bechaz, A., & Dai, D. W. (2024). Ethical views on AI translation; Haim, O., Orland-Barak, L., & Goldberg, T. (2022). Linguistic repertoires of bilingual teachers; Holmes, W., Bialik, M., & Fadel, C. (2019). Artificial intelligence in education; Hsiao, J.-C., & Chang, J. S. (2024). AI tools for EFL reading and writing; Huang, A. Y. Q., Lu, O. H. T., & Yang, S. J. H.(2023). AI-enabled personalised learning; Khan, M. A., Kurbonova, O., Abdullaev, D., Radie, A. H., & Basim, N.(2024). AI-assisted assessment for young learners; Kilian, K. A., Ventura, C. J., & Bailey, M. M. (2023). Risks and control of advanced AI; Lin, C.-H., Zhou, K., Li, L., & Sun, L. (2025). Generative AI and multimodal composition; Mancusi, L. (2024). AI-supported vocabulary learning; Nazari, N., Shabbir, M. S., & Setiawan, R. (2021). AI writing assistants in higher education; OECD (2023). Digital education outlook; Page, M. J., McKenzie, J. E., Bossuyt, P. M., et al. (2021). PRISMA 2020 guidelines; Pourdana, N., & Tavassoli, K. (2022). E-portfolio assessment and writing; Tai, T.-Y., & Chen, H. H.-J. (2024). Intelligent assistants and EFL speaking; Prestridge, S. (2024). Pedagogical beliefs and disruptive technologies; Woo, D. J., Wang, Y., Susanto, H., & Guo, K. (2023). Idea generation with AI tools; Woo, D. J., Wang, D., Guo, K., & Susanto, H. (2024). Teaching EFL writing with ChatGPT; Xia, Q., Chiu, T. K. F., Lee, M., Sanusi, I. T., Dai, Y., & Chai, C. S. (2022). Self-determination design for inclusive AI.