Abstract

Plurisemiotic discourses made by artificial intelligence constitute a novel type of communication that deserves our greater attention because of its ethical implications. This study specifically focuses on portrait-based AI art apps, technological tools that have become highly popular on social networks, used to generate images resembling human-made illustrations based on users’ photographs. Employing a discursive semiotics and sociosemiotics perspective, we discuss the circulation of these images on social networks, as well as the subsequent controversies regarding both the reproduction of stereotyped beauty canons and the images’ lack of originality. Through the analysis of one of the most popular portrait-based AI art apps, Lensa, we examine the role played by this form of technical mediation in the circulation of contemporary images. The examination of the narrative programs that characterize the propagation of AI art avatars reveals the value of partial and total repetition at the basis of the users’ interactions.