Using the method of jail(break)ing to study how the visualities of sensitive issues transform under the gaze of OpenAI ’s GPT 4o, we found that: -Jail(break)ing takes place when the prompts force the model to combine jailing (transforming or fine-tuning content to comply with content restrictions) and jailbreaking (attempting to bypass or circumvent these restrictions). – Image-to-text generation allows more space for controversy than text-to-image. – Visual outputs reveal issue-specific and shared transformation patterns for charged, ambiguous, or divisive artefacts. – These patterns include foregrounding the background or ‘dressing up’ (porn), imitative disambiguation (memes), pink-washing (protest), cartoonization/anonymization (war), and exaggeration of style (art).
By Alexandra Rosca, Elena Pilipets, Energy Ng, Esmée Colbourne, Marina Loureiro, Marloes Geboers, and Riccardo Ventura for Digital Methods Initiative on August 6, 2024