Taming the Element: How AI Is Integrating into the Educational Process Around the World
Artificial intelligence is gradually becoming an indispensable part of higher education. Both students and teachers use it to reduce the volume of routine tasks and expand their capabilities. The limitations and prospects of AI are discussed in the report ‘The Beginning of the End or a New Era? The Effects of Generative Artificial Intelligence (GAI) in Higher Education,’ published in the journal Modern Education Analytics, under the scientific supervision of HSE Academic Supervisor Yaroslav Kuzminov.
The authors themselves used AI in preparing the report. In particular, ChatGPT helped analyse articles, group research, structure the report, and improve the text's quality. AI is seen as an emergent innovation penetrating various segments of university activities, the report notes. The impact of artificial intelligence on education has grown exponentially since the emergence of GPT and similar models. There has also been a surge in interest in the topic from researchers since 2023. However, the main research issue is that AI in education is often viewed from a dual perspective: ‘good’ vs ‘bad,’ ‘for the student’ vs ‘for the teacher.’
‘The most popular research questions currently being actively discussed in scientific debates are: is AI dangerous for educational institutions as a phenomenon, and does it lead to degradation? Should the use of models be regulated in general, and where is the boundary between their useful and harmful applications? Will the teaching profession disappear? These questions are often posed separately, which is a mistake; education should always have two sides,’ noted Yaroslav Kuzminov.
The report considers these issues together in order to understand the cumulative effect and contribution of AI to education.
The authors of the report highlight the leading countries in terms of the number of publications on AI usage and the specifics of their thematic focus. For example, in the United States, most publications focus on the integration of AI into programming, medical education, STEM fields, and combating fraud. Chinese researchers are primarily interested in academic integrity, inclusive education, and the development of AI literacy. In the United Kingdom, studies focus on using AI in interdisciplinary research and creative writing. Australia's scientific interest lies in innovations in teaching, digital literacy, and support for students and teachers. Spain focuses on AI's role in supporting creative thinking and language learning. Researchers from India are interested in inclusive design, Malaysia in supporting self-learning, and Germany in interactive learning.
The prevalence of AI usage varies across countries, depending on the characteristics of their education systems, the digital skills of students and universities, and their proficiency in English. For instance, a study by authors from Hong Kong showed that over 60% of students used GAI for academic purposes, while the figure was just over 40% for the UAE, 85% for Lebanon, and about 21% for Russia.
The scientific papers widely describe how AI is used in education by both teachers and students. However, most works in this area are hypothetical or based on very localised practical experience. Fewer studies rely on large samples and advanced data analysis methods. Empirical works are often replaced by essays discussing the risks of using AI. A significant portion of the works includes recommendations for educational policy, and this share is much higher than average.
Many studies focus on the legality of AI use by students for completing academic tasks. While students are interested in using AI effectively and quickly to complete assignments and improve their AI skills, teachers strive to identify instances of its use and, in cases where it is explicitly prohibited, hold students accountable. Publications systematically discussing the real possibilities and, especially, limitations of AI are rare.
‘At the same time, the scientific debate surrounding AI in higher education often takes a speculative nature. Some authors have an incomplete or often even incorrect understanding of how AI tools work, confusing GAI with other forms of AI. Many of the recommendations offered in the literature are not evidence-based. They often do not provide practical tools for effectively implementing GAI in educational practice and improving the user experience,’ the report emphasises.
The potential of AI can only be fully realised with a balanced, ethically justified, and pedagogically sound approach to its implementation, believes Yaroslav Kuzminov.
‘There is a need to adapt curricula and assessment methods, teach students and teachers to use GAI responsibly, and develop effective measures against academic fraud (or, conversely, abandon forms of control where academic fraud using GAI is possible)—all of this requires further research and the introduction of new approaches,’ he noted. ‘Teachers and educational institutions must actively work on integrating these technologies to minimise risks and maximise the benefits of their application.’
The academic supervisor of HSE University believes that future research should critically focus on developing specific, practically applicable methodologies for integrating AI into the educational process, with proven effectiveness. There should also be more attention given to studying the long-term consequences of AI usage in education and developing ethical frameworks for its application.