ADAPTIVE AI TOOLS AS DIGITAL SCAFFOLDS FOR EFL WRITING
This article presents the findings of an applied study on the implementation of the “Dynamic Scaffolding” methodology, which integrates adaptive Artificial Intelligence (AI) tools into the teaching of academic writing in English for non-linguistic majors in Kazakhstan. The research addresses the challenge of significant heterogeneity in student language proficiency within a single classroom, a key issue in the context of the digital transformation of national education. A pilot study involving 55 students from META University was conducted, utilizing a diagnostic survey to assess student perceptions of AI tool usage across various aspects of writing: vocabulary, essay structure, grammar, and critical thinking. Results demonstrate the high perceived efficacy of AI as a tool for supporting structural, grammatical, and motivational aspects of learning, while simultaneously revealing its perceived limitations in developing higher-order cognitive skills as deep idea integration and critical thinking. The article proposes a structured pedagogical model for the conscious integration of AI, positioning it as a “digital scaffold” to personalize learning and free up instructor time for the development of complex analytical competencies.