Exploring the boundaries of authorship: a comparative analysis of AI-generated text and human academic writing in English literature

Abstract

As artificial intelligence (AI) increasingly permeates educational landscapes, its impact on academic writing has become a subject of intense scrutiny. This research delved into the nuanced dimensions of authorship and voice in academic writing, specifically focusing on the application of OpenAI’s ChatGPT. In this study, the research team compared and contrasted an essay written by one second-year English student for a course on English literature with a similar essay produced by ChatGPT. The current research also, tried to clarify whether artificial intelligence can satisfy the formal requirements of academic writing and maintain the distinctive voice inherent in human-authored content. The examination hinges on parameters such as assertiveness, self-identification, and authorial presence. Additionally, the researchers shed light on the challenges inherent in producing AI-generated academic text. While ChatGPT presented an ability to generate contextually relevant content, the results highlighted its need for support in guaranteeing factual accuracy and capturing the complex aspects of authorship that are common in human writing. Notably, when compared to human-generated text, the AI-generated text was deficient in terms of specificity, depth, and accurate source referencing. While AI has potential as an additional tool for academic writing, this study’s findings indicated that its current capabilities—particularly in producing academic text are limited, and remain constrained. This study emphasizes upon the imperative for continued refinement and augmentation of AI models to bridge the existing gaps in achieving a more seamless integration into the academic writing landscape.