Addressing fragmentation: A systematic review of risk and remedies associated with AI and English language teaching
Introduction: Artificial Intelligence (AI) shows significant potential in English Language Teaching (ELT); however, its use presents diverse challenges, highlighting a gap in systematic reviews that have not yet fully synthesised these issues and their associated remedies. Research Aim: Guided by PRISMA guidelines and Okoli’s (2015) four-phase framework, this study critically reviews literature on AI-related risks and associated remedial strategies in ELT, with a focus on higher education contexts. Evidence-based Facts: From an initial corpus of 350 studies, twenty peer-reviewed articles (2023–2024) were selected based on relevance, citation impact, and indexing in high-impact journals. A deductive thematic analysis informed by eleven predefined categories was conducted using reflexive thematic analysis (RTA) principles through MAXQDA 24. The synthesis indicates uneven conceptualisations of AI-related risks across studies. Ethical concerns, particularly academic integrity threats and risks linked to rapid AI adoption, dominate the literature and set the context for other challenges. These are closely followed by technical limitations and pedagogical shifts that involve insufficient digital literacy, teacher readiness, assessment-related constraints, and issues of access and equity. Learner responses vary in motivation, engagement, critical thinking, and emotional factors, although AI is generally associated with increased autonomy and participation. Skill development remains inconsistent across language domains. Proposed remedies cover pedagogical, ethical, technological, and governance dimensions, including training development, but remain largely conceptual with limited evidence of effectiveness. Summary: AI integration in ELT is still formative, characterised by fragmented evidence and limited synthesis of risks and responses. Stronger empirical inquiry and integrated frameworks are required for effective implementation.