Veggi, ManueleCatalano, Chiara EvaPescarin, SofiaCatalano, Chiara EvaParakkat, Amal Dev2025-05-092025-05-092025978-3-03868-271-4https://doi.org/10.2312/exw.20251058https://diglib.eg.org/handle/10.2312/exw20251058Generative AI has opened up new and largely unexplored opportunities to address the challenges of communicating cultural heritage through interactive experiences. This study explores the potential of text-to-image models to support the early stages of interactive media design for cultural heritage applications. We conducted a qualitative survey using four generative AI services (Stable Diffusion, Adobe Firefly, MidJourney, and DALL-E) to address a real design challenge in the cultural domain. While the study does not cover the full scope of traditional interactive media design workflows or provide a comprehensive performance evaluation, it highlights key benefits of generative AI in early design phases. The survey reveals that these systems boost creativity by introducing unexpected elements, they have the ability to help consolidate initial ideas and communicate them effectively to colleagues or stakeholders, and even help less experienced designers understand design requirements.Attribution 4.0 International LicenseCCS Concepts: Computing methodologies → Artificial intelligence; Graphics systems and interfaces; Human-centered computing → Empirical studies in interaction design; Applied computing → Fine artsComputing methodologies → Artificial intelligenceGraphics systems and interfacesHuman centered computing → Empirical studies in interaction designApplied computing → Fine arts10.2312/exw.202510589 pages