AI in Literary and Non-Literary Discourse: A Systematic Review of Human and Machine Authorship, and Style and Perception

Artificial intelligence has quickly invaded the realms of both creative and information-based writing, raising new questions about human originality, authorship and style. Despite its ability to produce writings that are coherent and stylistically varied, there are still concerns over the uniqueness and cultural neutrality of AI programs such as ChatGPT. This review covers significant recent advancements with artificial intelligence applications in both the literary and non-literary fields. It analyzes 35 recent studies contrasting authorship and creativity, or stylistic considerations and impressions, between human and AI texts. These studies range from poetic and fictional writing through essay, news article and academic publications. Six main topics emerged: Reconfiguring Authorship and Creativity in AI-Generated Literature; Disentangling Human and AI Authorship in informative and Creative Writing; Cultural Stereotyping, Stylistic Homogenization and Creative Constraints in AI-Generated Literary Content; Audience Perception, Misidentification and Heuristics in Informative Discourse; Educational Applications and Authorship Detection in AI-Generated vs. Human Texts; and Integration of AI into Special Research Fields such as Law, Criminology, Literature Reviews and Social Media. In particular, it was determined that though AI may have the ability to mirror humans on the basis of grammar, structure and at times a degree of creative literacy, it does not possess the capacity to address matters of deep emotional resonance and cultural context. However, at times humans cannot easily discern between texts written by AI and those authored by humans themselves, calling confidence and authority into question. Future studies could focus on how cultural background impacts the responses of people to AI-generated content, particularly in educational environments of colleges and universities.