Libro

The incorporation of Artificial Intelligence (AI) into second language writing instruction has introduced a dynamic paradigm where automated, instantaneous feedback serves as an interactive scaffolding mechanism. By targeting intermediate (CEFR B2) English learners, this research explores how digital corrective tools intervene during the drafting process to disrupt error fossilization and enhance syntactic precision. Rather than relying solely on traditional, delayed evaluations, the intervention allows students to engage in real-time hypothesis testing across complex linguistic structures. The empirical core of the study monitors shifts in five distinct grammatical dimensions—ranging from morphological agreement to structural punctuation—while simultaneously assessing the cognitive and emotional impact of this automated interaction on the student's writing experience. The outcomes reveal a dual advancement in both measurable linguistic performance and student self-regulation. Quantitatively, the post-intervention metrics show a clear upward trajectory in grammatical accuracy, demonstrating that targeted automated corrections successfully guide learners toward more precise sentence construction. Qualitatively, the data reflect a significant shift toward learner autonomy, as the immediate nature of the feedback removes conventional classroom anxieties and encourages independent revision habits. Although the scope of the study is constrained by its limited duration and small cohort, the findings provide a strong foundation for integrating hybrid teaching models. This sets the stage for future longitudinal studies to explore the long-term effects of AI on sustainable language development.