How AI is changing the field of education. Analysis
Higher education may abandon the traditional practice of writing theses due to the use of artificial intelligence (AI) models by students. The Sochi Institute of the RUDN University was the first to stop defending diplomas for bachelors studying in the field of Law. How access to generative models is changing the Russian education system and how it is handled in other countries is reviewed by Izvestia.
The role of AI in education
• Generative artificial intelligence is actively penetrating the educational sphere, which is facilitated by both weak regulation of the scope of AI and the rapid adaptation of children and adolescents to digital products. From 2022 to 2025, the share of generated text in students' graduation papers has more than quadrupled: from 9.9% to 42.3%. Most often, students turn to AI assistants to compose the introductory and final parts of the text: generative models help students overcome the "fear of a blank slate." AI is also used to compile a list of references — it helps to arrange the text in accordance with the requirements of the university, and at this stage many universities do not object to the use of an AI assistant.
• According to a study by the Higher School of Economics published on July 1, generative models are also used by teachers. 85% of Russian scientists use AI to compile scientific papers: it is most actively involved in social (89%), humanitarian (77%) and agricultural disciplines (76%), and the most popular field of application is the translation of texts into foreign languages. Students also often turn to an AI assistant when preparing projects and presentations.
• Instead of fighting the use of AI models in universities, the Russian Ministry of Education and Science suggests using the capabilities of artificial intelligence to review the educational approach and prepare the entire education system for a new generation of children born in 2025-2039, for whom the use of AI to perform various tasks will be the norm from birth. In particular, the option of returning oral exams to universities is being considered, which will exclude the participation of neural networks in the final assessment, and the student will have to answer the teacher's questions independently.
• AI tools can greatly facilitate the learning process for children with special needs. For example, if there are difficulties with reading, the AI assistant can translate the text of the textbook into an audio format. But some researchers note that the same effect can be achieved with digital tools that schools had before, so the question of the value of generative AI models for educational purposes and the boundaries of their application remain a matter of debate.
The risks of using AI
• In order for students to show real knowledge, universities use tools to identify plagiarism and generated content. But these measures sometimes backfire against those who write scientific papers on their own, without relying on AI. Generated text detectors can reject a work if it is written with errors or, conversely, demonstrates a wide vocabulary. Citations and introductory words may also be the reason why an original work may be considered generated.
• Due to the use of AI, the burden on teachers increases, who have to catch the mistakes of artificial assistants at the learning stage and independently check the work, since the "anti-plagiarism" systems produce erroneous results: this most often happens with works in the field of "Jurisprudence", where it is often necessary to quote laws, and texts in English, since some services mistakenly mark works created by a non-native speaker as generated. In fact, checking the text for AI participation turns into an exam for the teacher. It is he who can distinguish quotations from plagiarism and assess the quality of the knowledge gained by the student in terms of the depth of research — the work performed by AI is often quite superficial.
• AI tools can reduce the quality of cognitive skills of students and schoolchildren. A study by the Center for Strategic Corporate Forecasting at the Swiss Business School, published in 2025, showed that frequent use of AI tools reduces the ability to think critically. Moreover, young people who demonstrated lower critical thinking skills compared to older participants depend most on AI assistants. That is why the use of AI tools in schools and universities should be controlled.
• To prevent AI from harming, you need to be able to work with it. On June 29, the Russian Ministry of Education and Science announced the preparation of mandatory courses for students to work with artificial intelligence. Experiments with the use of AI tutors are already underway on online educational platforms. In fact, these are teacher assistants who can individually create homework assignments for students, help them complete them (when it is necessary to have a conversation on a given topic) and check the result. At the same time, the teacher plays a leading role in the educational process.
Foreign experience
• The United States and China are leaders in the development of generative AI models, and both countries consider it necessary to use AI tools in education. At the same time, American parents and the expert community oppose the introduction of artificial intelligence models in schools, since students who use AI show worse results than their peers who do not use generative models in their studies. Experts also see the increased use of AI models in schools as a threat to the psyche of participants in the educational process and the risks of violating confidentiality.
• In China, the introduction of AI into education is provided for by the state plan "A New Generation of Artificial Intelligence", adopted in 2017, and prepared by the "double reduction" reform — reducing the burden on schoolchildren and financial costs for parents who had to pay for additional education. The reform accelerated the use of digital tools in education. One of the main goals of AI in Chinese education is to provide an individual approach to each student. In conditions where one group can number 100 or more students, this problem cannot be solved without additional tools.
• AI tools form an entire educational ecosystem: sensors are used to analyze air quality, noise and light levels in classrooms, computer vision tools analyze the degree of student engagement. The AI assistant offers the teacher tasks to test knowledge and interactive graphs to make learning more visual, and a special system analyzes students' work to identify gaps in each student's knowledge and prepare a personalized program. Nevertheless, the Chinese model of AI application in education has drawbacks: it is criticized for "coaching" students on specific tasks to the detriment of creative thinking, the difference in equipment between elite and conventional universities increases digital inequality, and teachers complain of burnout — in 2023, 43% of teachers reported an increase in administrative burden.
Переведено сервисом «Яндекс Переводчик»