LEARNING GAINS AND PEDAGOGICAL IMPLICATIONS OF GENERATIVE AI IN ACADEMIC WRITING: A SYSTEMATIC REVIEW OF BS ENGLISH STUDENTS IN PAKISTAN

Abstract

The paper is a systematic literature review of the utilization of generative artificial intelligence (GenAI) and specifically ChatGPT in academic writing in higher education, focusing on BS English learners in Pakistan and other ESL/EFL settings. The analysis identified three important themes, including (1) academic writing learning gains, (2) ethical and academic integrity issues, and (3) implications in pedagogy and assessment. The results show that GenAI may positively impact grammar, coherence, sentence structure, and paragraph organization, especially when students actively use AI in brainstorming, revising, and reflecting. Nevertheless, there are still some issues of authorship, plagiarism, over-reliance, and the quality of outputs produced by AI. To adequately integrate GenAI, there must be a clear instruction of AI literacy, ethical guidance, and adjustment of assessment practices so that writing competence can be truly demonstrated. The paper identifies the conditional possibility of GenAI to aid academic writing, underlining that pedagogical intentionality, institutional frameworks, and responsible use are essential to the meaningful educational outcomes. Teaching and assessment implications and future research implications on the development of long-term writing in ESL/EFL settings are discussed.