This paper presents a research-driven framework to make conservation narratives visible in museums through AI-assisted content generation accessible to the public using. Technical conservation documentation is adapted for a general audience through ChatGPT, with all scripts and multimedia outputs reviewed and approved by a human expert. Synthesia (AI avatar video) and Clipchamp (AI voiceovers video) are evaluated for content creation, while QR codes and digital resources enable cost-effective visitor access. The framework is illustrated through the Archaeological Museum of Eleusis re-exhibition, demonstrating how AI tools, under full human supervision, can transform traditionally behind-the-scenes conservation work into engaging, educational, and participatory museum experiences.
Communicating conservation science to children requires transforming highly technical documentation into accessible, engaging, and scientifically reliable narratives. This paper presents a human in the loop framework that formalizes conservation documentation as epistemic ground truth and employs Generative AI exclusively for controlled narrative adaptation. The conservator retains full editorial authority, ensuring epistemic reliability and ethical compliance. The methodology is demonstrated through three comparative case studies implemented on freely-accessible/no-code platforms, Animaker, Genially, and Lumi Education/H5P, representing animated storytelling with puzzles, interactive flip card discovery, and quiz based reinforcement. The paper outlines future directions for evaluating the clarity and accessibility of the digital narratives within museum contexts. As part of an ongoing doctoral dissertation, the work outlines core research questions, methodological milestones, and future directions, including the release of an open repository to support reproducibility. The framework contributes a scalable, low-barrier model for responsible AI-assisted communication of conservation knowledge in museum contexts.