Children's narratives and generative artificial intelligence: ethical and pedagogical ghallenges in gender representation

Abstract

The advancement of Generative Artificial Intelligence (GenAI) has generated profound changes in the educational field, while also raising ethical and pedagogical questions, especially in the production of stories aimed at children. This study sought to critically analyze how GenAI contributes to the creation of children's literature, with special attention to the reproduction or contestation of gender stereotypes. The research adopted an exploratory qualitative approach, using Content Analysis to examine narratives produced by six chatbots (ChatGPT, Gemini Google, Jasper, Perplexity, Gamma and Story Spark). Each chatbot received three standardized prompts, generating a total of eighteen adventure narratives centered on scientific activities with insects. The findings reveal that GenAI can create stories that both reinforce and challenge stereotypes, depending on the chatbot and the prompt used. The consistent presence of child protagonists, engaged in investigations, data collection, and scientific record-keeping, stands out, promoting active participation and knowledge sharing. However, not all tools presented gender balance, and some maintained traditional stereotypes regarding scientists. The results emphasize the need for teacher and family mediation, encouraging critical reflections on diversity, representation, and gender roles in narratives.