Artificial intelligence in education: new horizons and challenges

Abstract

The spread of artificial intelligence in educational practice outpaces the development of pedagogical theory capable of describing the resulting changes in terms of didactics rather than mere informatization. The gap between the speed of technological adoption and the stability of the methodological framework gives rise to a situation in which the same tools are classified either as a factor that expands educational opportunities or as a source of cognitive and ethical risks. The aim of the study is to systematize the functions of artificial intelligence in education and to construct a classification that distinguishes the productive and risk-bearing effects of its use. A structural-functional approach, conceptual and categorial analysis, comparison of pedagogical concepts and content analysis of the normative basis of Russian education are applied. Four functional roles of artificial intelligence are identified – adaptive, diagnostic, organizational-managerial and generative – each associated with specific educational outcomes and attendant risks. It is established that the principal danger lies not in delegating operations to an algorithm as such, but in the shift of the learner’s agency position, which narrows the zone of independent intellectual effort. It is shown that the boundary between a horizon and a challenge runs not along the type of technology but along the mode of its didactic integration: one and the same generative model either reinforces or suppresses learning autonomy depending on the design of the task. No didactically neutral artificial intelligence tools are found, and the pedagogical effect is determined by the structure of the learning activity into which the tool is embedded.