INTEGRARE IL PROMPT ENGINEERING TRA LE ABILITÀ TRASVERSALI PER L’EDUCAZIONE LINGUISTICA: UNO STUDIO DI CASO

Integrating prompt engineering into the transversal skills for language education: a case study The implementation of Generative Artificial Intelligence (GAI), and particularly of Large Language Models (LLM), into the field of language education should not be limited to a purely instrumental use of AI-based technologies. Conversely, it requires the development of communicative, metacognitive and soft skills in order to foster a dynamic and synergistic human-AI collaboration. From this perspective, this paper presents the results of an experimental study conducted in a high school context and aimed at exploring the interactions between a group of students and ChatGPT during activities designed to enhance academic writing skills. The research focuses on the analysis of the prompt strategies adopted by participants to guide their interactions with the linguistic model by investigating how variations in the formulation and iterative refinment of prompts can influence the quality of ChatGPT’s feedback. Building on this analysis, the study proposes an operational model of transversal and digital competences, including mediation, critical thinking and problem solving skills, that learners and educators should develop to promote an effective and responsible use of LLMs for language teaching and learning.