Artificial intelligence in education: balancing personalized learning with cognitive and psychological risks

Abstract

The relevance of this study stems from the rapid integration of personalized educational technologies based on artificial intelligence (AI), which fundamentally reshape the interaction structure between learners, educators, and the learning environment. Despite the significant potential of individualized learning trajectories, there is growing concern about cognitive and psychological risks, including the decline of critical thinking, dependency on algorithmic prompts during task execution, and social isolation resulting from the automation of educational regulation. These issues necessitate a systematic analysis of the effects of such AI-based solutions, taking into account ethical, pedagogical, and technological factors. The article aims to identify the potential and limitations of using AI technologies in personalized learning in terms of supporting cognitive development, preserving learner autonomy, and ensuring psychological safety in education. Methodology. The study employs a comprehensive approach, combining a systematic review of scientific sources, a typological classification of personalized learning platforms, a structural-functional analysis of educational outcomes, and an examination of issues related to the implementation of AI in schools and higher education. Particular attention is paid to the transformation of pedagogical interactions, the cognitive effects of adaptive systems, and the ethical risks of automated decision-making. Results. The study finds that AI systems can enhance cognitive flexibility, metacognition, and learnersʼ self-organization skills. However, in the absence of appropriate pedagogical moderation, they may lead to reduced independence, disruption of learning balance, and loss of motivation. The structural barriers to implementing personalized AI solutions are identified, including technological limitations, algorithmic opacity, poor integration into local educational contexts, and insufficient methodological readiness of educators. Conclusions. The effective integration of AI in education requires a flexible, pedagogy-driven design, transparent algorithmic architecture, and a balanced combination of individual and social learning components. AI should serve not as a replacement for teachers, but as a tool for fostering thinking, reflection, and responsible learner autonomy. Future research should focus on empirical studies of the long-term impact of AI platforms on learning motivation, self-regulation, psychological well-being, and inclusivity in education, as well as on the development of ethical protocols for algorithmic interaction in the learning process.