A escrita como processo mediado e o papel da inteligência artificial generativa no apoio à produção de textos

Abstract

The widespread accessibility of Generative Artificial Intelligence (GenAI) presents significant challenges to teachers, particularly within the context of Higher Education. According to a recent systematic literature review focused on higher education teachers and students, the latter group appears more receptive to these technologies, acknowledging both the opportunities and challenges they entail. In contrast, teachers tend to exhibit greater skepticism, primarily due to concerns about potential risks to ethical standards, assessment integrity, and issues related to privacy and security (Ribeiras, Dorotea, & Esteves, 2025). In Higher Education, the use of Large Language Models (LLMs) may exert a notable influence on students' written production, a domain that remains essential for academic performance due to its transversal nature across most learning outcomes. This study aims to present how, within the course unit XXX, part of a Master's degree program for future teachers of the 1^st and 2^nd cycles of Primary Education, GenAI tools were integrated to enhance students’ writing proficiency. These tools provide personalized support and enrich the learning process by enabling tailored pedagogical strategies. The adaptive, individualized interaction fostered by such technologies also facilitates the implementation of collaborative activities, such as peer writing, under the guidance of the instructor. While GenAI can support multiple stages of the writing process, current literature suggests that its most effective use lies in the revision phase. Specifically, the act of crafting prompts to generate model texts can help students organize and structure their ideas more effectively, while also providing immediate and individualized feedback and feedforward. This contributes to the development of essential skills such as idea restructuring and lexical refinement, which are critical for producing reflective and critical writing. Furthermore, comparing student-generated texts with those produced by GenAI encourages critical thinking and self-regulation, enabling students to identify gaps and areas for improvement in their writing. With appropriate pedagogical supervision and grounded in a process-oriented approach to writing (Flower & Hayes, 1981), GenAI can indeed serve as a powerful tool to foster writing competence, critical thinking, and creativity, while being ethically and effectively integrated into academic contexts.