Images made by Artificial Intelligence:

ethical dilemmas and algorithmic biases

Abstract

Generative Artificial Intelligence has triggered a revolution in various spheres of society (Boden, 2020), creating new and authentic content. Although fascinating, this new way of making raised questions regarding authorship and perspectives on creativity and autonomy (Santaella, 2023), the reframing of the past, and the imminent fear of technology. Through a bibliographical review, this article explores ethical dilemmas (Coeckelbergh, 2024) and algorithmic biases (Faustino & Lippold, 2023) in creating images through AI platforms. In addition to this understanding, the work proposes to categorize, based on theories about memory (Benjamin, 2012; Halbwachs, 1990; Huyssen, 2000), the new forms of remembrance, through imagery creations that refer to the past. This path will use Content Analysis as methodological support (Krippendorff, 2004; Sanglard, 2017; Sampaio and Lycarião, 2021). The intention is to understand how memory is portrayed, reconstructed, or given new meaning by image generation tools with Artificial Intelligence and to understand the dilemmas emerging from this new way of remembering.