Generative AI as an English Writing Aid: Thai University Students’ Perceptions and Experiences with ChatGPT and Gemini

Abstract

Background and Objectives: The emergence of AI-driven writing assistants has sparked discussions on the potential benefits and drawbacks of these tools in higher education. While AI tools can enhance writing skills, provide instant feedback, and facilitate brainstorming, concerns persist regarding academic integrity, ethical considerations, and over-reliance on AI. This study aimed to explore Thai university students' perceptions and experiences with generative AI tools—ChatGPT and Gemini—within the context of academic writing in English. By examining these students' subjective experiences, this research also sought to better understand the benefits and risks associated with AI-driven writing assistants in higher education, particularly in Thailand’s unique academic environment.

Methodology: A qualitative approach was employed using semi-structured interviews with 12 Thai university students selected through purposive sampling. The study applied thematic analysis to identify key patterns and insights from the participants' responses. NVivo software was used for data organization. The study was grounded in the constructivist paradigm, emphasizing students' subjective experiences and contextual understanding of AI usage in academic writing.

Main Results: Thai university students view ChatGPT and Gemini as valuable aids for academic writing, particularly in brainstorming, structuring ideas, and improving grammar. AI-assisted feedback boosted confidence and writing quality, but concerns about over-reliance, academic integrity, and ethical considerations were prominent. Students employed strategies like paraphrasing and cross-referencing AI-generated content to ensure originality. While AI enhanced language learning through real-time feedback, some feared it might lead to superficial learning and reduced engagement in skill development. The findings underscore the need for clear academic guidelines to help students balance AI use with independent learning.

Discussions: This study highlights the dual nature of AI integration in academic writing—offering both significant advantages and potential risks. Students demonstrated a pragmatic approach, leveraging AI for efficiency while maintaining an awareness of ethical considerations. Their cautious engagement suggests that AI is seen as a supplementary tool rather than a complete replacement for traditional writing and learning methods. The findings align with broader discussions on responsible AI use in education, emphasizing the importance of balanced and mindful engagement with AI technologies. The study also underscores the importance of institutions developing clear guidelines on AI usage, as well as offering digital literacy programs that can help students navigate the ethical and practical aspects of AI integration. These efforts could ensure that AI tools enhance the learning experience without compromising academic integrity or the development of essential writing skills.

Conclusions: This study provides valuable empirical insights into how Thai university students perceive and utilize AI tools, particularly ChatGPT and Gemini, in academic writing. AI tools were found to significantly support writing quality and language development, yet concerns over over-reliance, ethics, and integrity remain. The findings stress the necessity of institutional policies and structured guidance to foster responsible AI use. Future research should include cross-cultural comparisons to examine variations in AI adoption within higher education. Longitudinal studies could assess AI’s long-term impact on writing proficiency, and targeted interventions should be developed to promote balanced and ethical AI integration in academic contexts.