ALGORITHMS VS. AUTONOMY: THE RISKS OF DEPENDENCE ON AI IN THE CRITICAL TRAINING OF STUDENTS

Abstract

This article critically examines the impacts of artificial intelligence (AI) on education, highlighting how algorithmic mediation can compromise students’ intellectual autonomy and critical thinking. The analysis reveals that adaptive platforms, automated assessment systems, and generative tools, while promising efficiency and personalization, often reduce learning to standardized processes, limiting the capacity for autonomous judgment and the construction of meaningful knowledge. The erosion of autonomy manifests itself in student passivity induced by predefined learning paths, while dependence on generative AI atrophies original argumentation. Furthermore, algorithms reproduce cultural biases and prioritize quantifiable metrics over qualitative dimensions of education. As alternatives, we propose active teacher mediation, where the teacher acts as a critical filter of algorithmic content, and hybrid models that preserve student agency. We also defend the need for ethical regulation, with transparency in algorithmic criteria and protection of educational data. The paper concludes that AI in education requires a delicate balance: if adopted uncritically, it can reinforce inequalities and impoverish human development; if integrated with solid pedagogical foundations, it can broaden access without sacrificing intellectual depth. Future research should investigate the long-term cognitive effects and develop truly inclusive systems.

DOI: https://doi.org/10.56238/sevened2025.026-014