The use of generative neural network assistants in the educational process: threats and new opportunities

Abstract

Introduction . Increasingly, teachers and students are using generative neural network assistants that transform the educational environment. Purpose setting . The article discusses AI services: GigaChat, YandexGPT, ChatGPT, DeepSeek. The purpose of our research is to identify new opportunities and threats related to their use in education. Methodology and methods of the study . The following methods were used: monographic, system analysis, statistical, including surveys, summary, grouping, calculation of relative values. The authors analyzed the features of using the services in question: the tariff, the target audience, and the availability on the territory of the Russian Federation. Two statistical studies have been conducted in the form of surveys on the use of generative neural network assistants by teachers and students of four Russian universities. Results . The functions of generative neural network assistants that are relevant for education are listed, which are more often used by students than by teachers. This is confirmed by the conducted statistical study: only 47.96 % of teachers in 2025. Generative neural network assistants were used, while first – year undergraduate students accounted for 82.05 %, and senior and graduate students accounted for 100 %. Our second study shows problems in recognizing the generated text. The authors see the following threats to the use of generative neural network assistants: the presence of «hallucinations» in services that are not always recognized by students; insufficient assimilation of the material studied; decreased emotional intelligence, the ability to learn, and develop critical thinking. Although the analysis of publications shows that some of the authors disagree with this and advocate the cybernization of consciousness, predicting that neural interfaces will eliminate these threats. Another consequence of using generative neural network assistants is a reduction in cognitive load. Scientists disagree on whether this is a good thing or a bad thing. On the one hand, this will help to avoid professional burnout and overload in humans, and on the other hand, it will weaken mental abilities, passive consumption of information, and degradation of existing skills. Generative neural network assistants are changing the educational environment, and only teachers with AI competencies can minimize the risks of such a transformation.

Ähnliche Arbeiten