THE ROLE OF AI-POWERED FEEDBACK IN IMPROVING ESL WRITING AT PAKISTANI UNIVERSITIES

Abstract

The rapid advancement of artificial intelligence (AI) has transformed educational practices worldwide, particularly in the field of language learning and writing instruction. AI-powered writing tools such as ChatGPT, Grammarly, QuillBot, and Gemini have increasingly been adopted by students to support various aspects of academic writing, including grammar correction, vocabulary enhancement, organization, and revision. Despite the growing popularity of these technologies, limited research has explored their role in ESL writing development within the Pakistani higher education context. This study investigated the role of AI-powered feedback in improving ESL writing among university students in Pakistan. Adopting a qualitative multiple-case study design, data were collected through semi-structured interviews, reflective learning journals, and writing samples from undergraduate students who regularly utilized AI-assisted writing tools. The collected data were analyzed using Braun and Clarke’s thematic analysis framework. The findings revealed that AI-powered feedback contributed positively to students’ perceived improvement in linguistic accuracy, academic vocabulary, coherence, organization, learner autonomy, and writing confidence. Participants reported that immediate access to AI-generated feedback encouraged active revision practices and supported independent learning. However, the study also identified concerns regarding the accuracy and contextual appropriateness of AI-generated suggestions, the potential for overreliance on technology, and ethical issues related to academic integrity. Furthermore, students expressed a strong preference for a balanced approach that combines AI-generated feedback with teacher feedback, recognizing the complementary strengths of both forms of support. The study concludes that AI-powered feedback has considerable potential to enhance ESL writing instruction in Pakistani universities when implemented responsibly and pedagogically. The findings contribute to the growing literature on technology-enhanced language learning and provide practical implications for educators, institutions, and policymakers seeking to integrate AI technologies into academic writing instruction.