BETWEEN HUMANS AND ALGORITHMS: authorship, style and enunciation in times of artificial intelligence

Abstract

: This article investigates the implications of using artificial intelligence (AI), especially ChatGPT, in the production of academic texts, in light of the Bakhtin Circle's theory. The analysis focuses on the concepts of utterance, intonation, style, and authorship, seeking to understand how these categories manifest in the discursive dynamic between humans and AI. The hypothesis is that ChatGPT functions as an implicit co-author, configuring a new form of verbal interaction marked by asymmetric alterity. The methodology adopted is qualitative, based on Bakhtinian discourse analysis of an article generated by AI from predefined prompts. The results reveal that texts produced with the help of ChatGPT participate in a discursive chain, incorporating simulated intonations and hybrid styles, in which the human author adjusts and guides the discourse according to their purposes. In this context, authorship emerges as a co-construction, requiring new ethical and epistemological considerations about originality and responsibility in academic creation. The study contributes to a critical understanding of the role of AI in education, proposing a dialogic perspective on language and knowledge production in the digital age.