Artificial intelligence in instructional design: preliminary findings on Italian teachers’ perceptions

Artificial Intelligence (AI) is reshaping education, generating both opportunities and tensions: alongside promises of personalized learning, expanded access to resources, and automation of routine tasks, concerns persist regarding algorithmic decision-making, transparency, surveillance, and the possible weakening of teachers’ mediating and relational role. The rapid diffusion of generative AI has intensified these debates, raising questions about knowledge authenticity and the aims of schooling. This paper reports preliminary findings from a descriptive quantitative survey administered to 460 Italian in-service teachers via I.D.E.A. Questionnaire, investigating familiarity, frequency and domains of use, and tool-specific adoption. Results indicate selective, incremental integration: teachers mainly rely on widely familiar tools (e.g., ChatGPT, Google Translate, Canva, Wordwall, Kahoot!), primarily for text-related tasks (summarising, rephrasing, translating) and material preparation, while advanced uses (simulations, data analysis) remain marginal. Assessment-related functions are rarely delegated to AI, suggesting protection of professional responsibility. Cluster analyses identify distinct adoption profiles (non-users, selective users, heavy users), pointing to differentiated needs for support and targeted professional development within a pedagogical and regulatory framework.