Caterina Morelli

Digitization of cultural heritage expands access possibilities but does not guarantee inclusion by itself. The article analyzes the role of alt-text in the accessibility of digital visual heritage and the potential of generative artificial intelligence to produce large-scale alternative descriptions. Through recent experiments in the cultural sector, it highlights opportunities and limitations, including factual errors, biases, and simplifications. AI emerges as a support tool, to be integrated with domain expertise, accessibility, AI literacy, and involvement of visually impaired people.