Neural networks create new foundations of education in Russian schools and universities

Artificial intelligence and neural networks were fantastic yesterday, and today they have become a slightly frightening part of everyday life. ChatGPT, SberIndex, YandexGPT - these spells have not been heard in school and university walls before. Now they can be heard more often than the rules of the Russian language, Fermat's theorem, or the laws of thermodynamics. Neural networks quickly penetrated into the field of education. Schoolchildren and students energetically use neural networks to study, write course and abstracts, generate notes, and decipher complex lectures. Teachers - for preparing and checking lessons.

Is the Ministry of Education trying to control this lightning-fast transformation? In 89 regions of Russia, schools are testing neural networks to control exams and analyze student performance. But, so far, there is no systemic policy regarding the role of AI in the system. The Ministry of Education and the Ministry of Education and Science act with an eye, usually verified by prohibitions and recommendations to teachers. However, there are non-systemic attempts to "ride the wave," and quite numerous.

Dozens of educational projects have been launched in Russia in which neural networks help training: from automatically checking homework to creating personal training plans. Basically, these innovations are tested in "pilot", developed regions.

In Moscow, Tatarstan, Novosibirsk, AI-based platforms help teachers to individually approach each student. Online schools such as Foxford are implementing AI algorithms for teaching guidance. At the Moscow Electronic School, AI-based services analyze the knowledge of students and select tasks for each of them. Some universities in Moscow (in particular, HSE, Bauman Moscow State Technical University) teach AI courses and at the same time develop their own models for monitoring academic performance.

At Kazan Federal University, teachers are also actively experimenting with AI: generating tests, adapting content to the student level, creating digital avatars. Some schools in Tatarstan have similar courses for children: they teach them to work with neural networks, create their own projects - from cartoons to games.

I must say that these examples do not at all reflect the true depth of penetration of neural networks into the educational process. According to polls by VTsIOM and SberEducation, more than 60% of high school students have used ChatGPT and AI analogues at least once to write essays and solve problems. In general, about 64% of young people regularly use neural networks to search for information, 23% of schoolchildren turn to AI while preparing for exams, and 17% - to read a simple explanation of difficult topics. This means that AI technologies have already taken root in schools and universities, and in the future their importance will only increase.

There are pluses to this that cannot be ignored. For example, GPT-based adaptive platforms generate tasks based on the student's current level of knowledge and ability. Neural networks help schoolchildren with disabilities and developmental disabilities. AI can voice texts, or translate gestures into text, neural networks simplify formulations that are difficult to understand. Finally, it should be borne in mind that for children of the generation of smartphones and computers, interaction with neural networks is a natural, familiar format. It makes learning more attractive.

AI removes most of the routine load from school teachers and university teachers: there is no need to manually pore over tests and check errors in hundreds of tasks. This gives time for creative, lively work with children.

Now about the negative consequences of the widespread introduction of neural networks into the educational process. Along with convenience, new risks and problems appear that are important to see in time.

A neural network helps to explain complex topics, can instantly find the information you need, competently generate text, create a picture, or video. On the one hand, this is a technological leap, on the other, it is a modern form of "cheating," which carries the risk of reducing the quality of knowledge. This is the first serious problem.

The second problem: AI is still a fine-tuning tool, but it is initially configured by people who may not always have good intentions. Recently, neural networks have repeatedly come across the issuance of inaccurate or absolutely false information. AI can pass off as "face value" and specially created misinformation that will look extremely plausible.

Obviously, neural network fakes can affect the worldview and safety of schoolchildren. In parallel with the basic course of work with AI, the younger generation needs to teach courses on the safe use of these technologies. It is required to teach children to be critical of photo videos and text information from neural networks, to be able to distinguish reality from fake. Otherwise, they will easily become victims of disinformation, or fraudulent manipulators. At the same time, adolescents themselves must learn to ethically use neural networks in their studies and creativity.

In a word, work for the Ministry of Education is a rough land. The adaptation of new technologies into the educational process requires not even reforms, but revolutionary transformations. But the Russian education system is still hesitant at a fork in the road. Here is what has been done over the years of explosive development of AI technologies:

In 2023, the Ministry of Education recommended "limiting the use of AI" in schools when performing homework.

In 2024, a pilot project was launched to teach teachers digital skills, including the basics of working with neural networks. Digital literacy programs have been launched in Moscow and St. Petersburg, Tatarstan, but other regions remain outside the framework.

In 2025, the "Concept of Digital Transformation of Education" spelled out a clause on the study of AI within the framework of the subject "Informatics." But this item is still optional.

Here, in fact, is the entire list of "directives" from the Ministry of Education. Meanwhile, the ministry needs to work proactively, proactively. It is necessary to integrate AI into the educational process consciously. Do not hide from the onset of neural networks, do not prohibit, but correctly and accurately direct. With the full understanding that this is a new inevitable stage in the development of the education system.

Already now, offhand, we can list a number of necessary acts that will contribute to the mutual adaptation of the digital world and classical school and university education:

Creating a legislative framework on ethics and restrictions on the use of AI in education.

Publication of uniform guidelines for schools.

Systems analytics and transparent AI risk assessment.

Retraining teachers for the effective use of neural networks.

This work will have to involve not only specialists from the ministry, teachers, but also professionals from the IT sphere, scientists, sociologists and representatives of public organizations. This approach will help make the school curriculum as modern as possible, interesting and useful for a generation of schoolchildren in the 2020s. The choice here is generally small: either the system will adapt and create new learning rules, or it will begin to slip and lag behind the students themselves, who have already lived in the neural network world for the last 2-3 years.