Generative AI as an interactive learning environment in early childhood: preschool children’s emerging AI literacy as socially distributed interpretive work in collaborative storytelling

Generative artificial intelligence (AI) is increasingly being used in early childhood classrooms, yet there is little understanding of how AI-related interpretive practices unfold in teacher-led activities. This qualitative case study examined whole-class AI-assisted storytelling in a public preschool in southeastern Türkiye. The corpus consists of 17 Turkish transcript files from the episodes (involving 22 children aged 5–6 years), along with available video and image records. In the individual episodes that were examined, the children treated the AI as an image producer, offered agentive explanations for unexpected outputs, and on occasion, linked features of the prompt to the resulting outputs. The teacher wrote all the prompts and was responsible for the majority of the speaker-labelled dialogue; her actions in many instances opened up certain lines of inquiry, in others, provided causal explanations, and, on a number of occasions, closed down or redirected the children’s lines of inquiry. Request fidelity, story suitability, and plausibility are thus presented as coexisting analytic orientations. Our study indicates that early AI literacy can be observed as socially distributed interpretive work within a deliberately structured interactive learning context.