Trustworthy Digital Mediation for Renaissance Heritage: Authenticity, Bias, and Cognitive Security in AI Assisted Storytelling

Digital technologies and AI are rapidly transforming cultural mediation, enabling scalable cataloguing, personalized interpretation, and immersive experiences that bring historical content to new audiences. In the MARY4ALL context—where Renaissance architecture, arts and music associated with Mary of Hungary are digitally re-activated through multisensory installations and digital storytelling—the question is no longer whether AI can enhance access, but under which conditions it can remain trustworthy, inclusive and sustainable. This paper proposes a practical framework for “trustworthy digital mediation” that addresses three intertwined dimensions: (i) authenticity and provenance of AI-assisted reconstructions (including disclosure of synthetic elements, documentation of sources, and uncertainty communication); (ii) bias and representational integrity in narrative choices and personalization (e.g., whose heritage is foregrounded, which voices are omitted, and how recommendation logics may steer collective memory); and (iii) cognitive security risks affecting public history, such as narrative manipulation, deepfake-enabled confusion, and attention-driven polarization when heritage content circulates through platform ecosystems. Building on research on information/cognitive warfare, human rights and governance of emerging technologies, the paper translates security and rights-based concerns into museum-ready safeguards and governance practices—lightweight enough for heritage institutions, yet robust enough to support accountability. The contribution concludes with a concise set of implementable recommendations (checklists and “red flags”) for curators, digital producers and cultural-route managers, aimed at strengthening public trust while preserving cultural integrity and ensuring socially and environmentally responsible innovation.