Riflessività sintetica Feedback di secondo ordine nell'intreccio tra contenuti dell'Intelligenza Artificiale Generativa e consapevolezza sociale

This contribution introduces the concept of synthetic reflexivity to critically analyze the interactions between Generative Artificial Intelligence and social awareness. Through a case study based on focus groups with university students and the use of images produced by Stable Diffusion, the article explores how cultural, gender, and ethnic biases emerge in AI-generated content and how they are recognized, corrected, and reinterpreted by users. Synthetic reflexivity is framed as a second-order observation practice that involves not only the analysis of outputs, but also the examination of the assumptions through which users make sense of the algorithmic «black box». The findings highlight how generative models do not merely reflect reality but actively contribute to constructing it, amplifying dominant narratives. Synthetic reflexivity thus becomes a critical tool for cultural studies to deconstruct stereotypes, problematize representations, and foster more inclusive forms of collective imagination.