English for Academic Purposes (EAP) in the Age of Generative AI: Redefining Academic Literacy and Writing Practices in Higher Education

Abstract

University students are adapting to new ways of planning, composing, revising and evaluating academic texts with generative artificial intelligence (GenAI). The core challenge in English for Academic Purposes (EAP) goes beyond the question of whether it is acceptable to use writing assistance from AI. The main problem is how to define the ability to write academically when fluent text can be produced at will? To this end, the article draws on a systematic review of the literature on EAP, academic writing, AI literacy, assessment and academic integrity (with an associated structured, PRISMA-guided methodology), preceding a conceptual synthesis. The results show that, for academic writing, competencies at multiple levels are needed: linguistic/rhetorical mastery; critical judgements on the potential contributions of AI-generated content to knowledge and regulation in relation to academia; wise integration of AI into the process of writing (rather than a final touch); and judgement around the ethical implications of doing so. Results also suggest a requirement for new ways to assess learning that embody the iterative nature of learning, such as staged drafts, reflective commentary, oral explanations and transparent records of using an AI. Instead of undermining teachers, GenAI re-constitutes educators as exemplars of judgement within disciplines, source evaluators, facilitators of ethical analysis and critical readers/writers interrogating AI-generated text. Based on these insights, the article presents the Academic-AI Literacy Integration (AALI) model which bridges academic literacy, AI literacy, and responsible digital pedagogy. It does so by framing responsible human AI collaboration not as authorship — or even learning, but rather as a practice that must comply 34 with disciplinary standards, evidentiary requirements and human evaluation.