AI AND RENAISSANCE POETRY How Prompt Language Shapes Literary Translation

AbstractThis conference companion presents an extended comparative analysis of AI-generated literary translation in Renaissance Italian poetry. Focusing on Petrarch's sonnet “Erano i capei d’oro a l’aura sparsi”, the research compares two AI-generated translations produced through Italian and English prompts against Joseph Tusiani’s human translation as the gold standard. The analysis The comparison is articulated along four observational axes including fidelity, rhythm, stylistic adaptation and poetic structure, and highlights how prompt formulation may influence the translation strategies in Large Language Models. The comparison suggests that Italian prompts tend to favour more literal renderings, while English prompts encourage more adaptive and stylistically creative outputs. The study further reflects on the current limitations of AI systems when dealing with densely structured poetic language and on the continuing importance of human interpretative expertise, empathy and sensitivity in literary translation.