Simulacrum in algorithmic social media feeds: perception of truth in the age of artificial visuality

Abstract

In today’s digital environment, the volume of visual imagery generated by artificial intelligence and disseminated through algorithmic social media feeds is steadily increasing. This transformation of media content raises the critical issue of shifting focus from documentary truth to emotionally plausible simulation, which significantly affects users’ cognitive perception of truth. The purpose of this study is to analyze the functioning of visual simulacra in personalized feeds and to examine their impact on the formation of perceptions of information authenticity. The methodological framework combines an interdisciplinary approach integrating J. Baudrillard’s theory of simulacra, visual semiotics, critical discourse analysis, and digital ethnography. The empirical basis consists of Facebook posts from 2024–2025 containing photorealistic images with identifiable features of artificial generation. Methods applied include visual and content analysis as well as comparative analysis of authentic and AI-generated imagery. The findings reveal that images without real-world prototypes, yet constructed according to the principle of emotional recognizability, are perceived as authentic due to narrative framing, orthographic imperfections, selective or absent use of hashtags, and aesthetic stylization. Algorithmic amplification enhances the virality of such materials while diminishing the audience’s critical vigilance. It has been proven that: 1) visual simulacra function as elements of hyperreality, replacing documentary evidence with affective reactions, especially in conditions of social tension or war; 2) the effectiveness of simulacra in the digital environment is due to their multimodal nature: the visual and verbal form a single communicative construct, where the text acts as a framework for interpretation and sets the trajectory of perception of what is depicted; 3) accompanying texts are formed according to two basic models (a character’s address to the user or a conditional narrator’s statement), each of which has a specific impact on the mechanisms of trust and emotional engagement.

The research novelty lies in integrating the concept of simulacra with the analysis of algorithmic content promotion mechanisms, providing deeper insights into the dynamics of media truth construction in the age of artificial visuality. The practical value of the study lies in identifying mechanisms of visual manipulation and developing recommendations for enhancing digital visual literacy.