Modern education is confronted with a fundamental contradiction between traditional teaching methods and the cognitive habits of the new generation of learners. As Morozov points out, the “digitalisation of students’ consciousness” means that visual and interactive content has become their natural environment, whereas traditional lessons, which rely on text and the teacher’s monologue, are losing their effectiveness [7]. Furthermore, in classes with mixed levels of preparation, the problem of differentiation is acute: a teacher cannot simultaneously design twenty‑five individual educational trajectories [7]. At the same time, as Zabelina, Pinchuk and Gritskova emphasise, “the development of modern information technologies, especially artificial intelligence, is significantly transforming the field of education and creating a need to transform traditional teaching approaches” [8]. AI can improve learning effectiveness through personalisation, adaptive programmes, objective knowledge assessment and increased student motivation [8]. The aim of this article is to identify, on the basis of a synthesis of practical experience and research findings reported in the current pedagogical literature, the main directions, methodological techniques and conditions for the effective use of AI technologies in school history lessons.

Understanding the current state of AI in education is impossible without looking at its history. As Zabelina, Pinchuk and Gritskova note, the first AI‑based educational systems appeared as early as the 1960s and 1970s. Among them was PLATO (Programmed Logic for Automatic Teaching Operations), developed at the University of Illinois, which included electronic textbooks, tests and even elements of online communication [8]. In the 1970s, the SCHOLAR system for teacher‑learner interaction was created, as well as GUIDON, which helped medical students make diagnoses. Despite offering individualisation and rapid feedback, these systems suffered from a lack of genuine understanding of learners’ needs and from high development costs [8]. In the 1990s, more advanced systems emerged: ALEKS (University of California), which identified gaps in knowledge and suggested appropriate materials, and ActiveMath for mathematics learning [8]. From the 2000s to the present day, systems such as Knewton and Dreambox Learning have focused on mass adaptive learning, performance analytics, the use of chatbots and the generation of learning materials [8]. Koroleva also observes that “in recent years there has been growing interest in using AI to create interactive learning platforms that can adapt to the individual needs of students” [4]. She identifies categories such as adaptive learning systems, chatbots, interactive simulations and big data analytics [4].

One of the most promising directions is the use of neural networks to generate visual images of historical eras, events, figures and architecture. Morozov notes the effectiveness of the YandexGPT and Kandinsky 3.1 neural networks for these purposes [7]. He has developed a visualisation algorithm that includes defining the lesson’s topic and objectives, selecting terms and concepts, determining the most effective mode of visual presentation, directly creating the visual materials, and then having students reflect on their work [7]. Masimzade emphasises the use of virtual and augmented reality (VR and AR): when studying the Great Geographical Discoveries, students “live” follow the pioneers along virtual routes, which “enables students not only to visualise history but also to develop spatial thinking and analytical skills” [5]. According to her surveys, the introduction of AR technologies increased interest in history lessons by 30 % [5]. Fitzner adds that “neural networks can generate realistic virtual models of historical events, allowing students to ‘visit’ the past, see it with their own eyes and feel the atmosphere of the era” [1]. Koroleva gives an example: “students can ‘visit’ historical sites using virtual reality, which makes lessons more engaging and memorable” [4]. Mishenina offers original, even provocative visualisation techniques: generating images of historical figures (Lenin, Catherine II, Peter I) as Barbie and Ken using the BaiRBIE.me neural network, which helps engage students in formulating the lesson’s topic [6]. She also uses the Midjourney neural network to create images of cities as people (Pskov and Yaroslavl are shown as elders, reflecting their ancient history) [6]. Another technique is “bringing portraits to life” using the Deep Nostalgia neural network from MyHeritage, which produces a video with facial animation; this can even be used during a physical‑activity break [6].

The personalisation and differentiation of learning is a crucial task. Morozov achieves differentiation by using AI chatbots (e.g., YandexGPT) to create variant exercises of the same type but at different levels of difficulty; to create personalised reading texts that take into account the student’s interests and reading‑literacy level; and to implement the “flipped classroom” model by having the AI generate individual reference notes [7]. Fitzner stresses that “AI is capable of analysing large amounts of information, identifying patterns and offering an individual approach to each student” [1]. Platforms such as Khan Academy or Learnosity already use algorithms that adjust the difficulty of tasks to the student’s level. “In history lessons, this can be seen in interactive tests that automatically generate questions of varying degrees of difficulty based on the student’s previous answers” [1]. Masimzade has created a database of tests and multi‑level assignments that allows each student to receive tasks corresponding to their level of preparation. AI is also used to analyse academic performance: “at the beginning of each topic, students take an online test, after which I configure a set of tasks for in‑depth study of the material. As a result, the adaptive approach made it possible to increase the average history grade by 15 % over two years” [5].

Developing critical thinking is another key area where AI can be applied. Klimentyev (under the supervision of Kuzmenko) carried out experimental work on developing critical thinking in tenth‑grade students using the YandexGPT neural network. Using the critical‑thinking assessment test developed by Gushchin and Ilyasov, he found that students showed the lowest results in the sections “Ability to evaluate sequences of inferences” (tasks 5‑6) and “Ability to detect errors connected with the uncertainty and ambiguity of expressions and terms” (tasks 13‑14) [3]. Special techniques were developed to address these weaknesses. The “Logical Chain” technique: students are given a set of historical facts generated by the teacher in YandexGPT in a mixed order; they must arrange them in a logically justified sequence, establish cause‑and‑effect relationships and formulate a conclusion [3]. The “Ambiguous Terms” technique: students receive a historical text (for example, an excerpt from the 1936 Stalin Constitution) containing terms with double meanings; they must identify the ambiguity of these concepts, explain why it arises, and demonstrate the discrepancies between the stated aims and actual historical practice [3]. Masimzade applies Edward de Bono’s “Six Thinking Hats” technique to develop critical thinking and teaches students to analyse media, recognise fake news and verify facts [5]. She also introduces online debates (e.g., on “Ethical aspects of artificial intelligence”), which develop students’ ability to argue and defend their position [5]. Fitzner likewise emphasises that “students can use AI to generate different versions of historical events and analyse their credibility and validity, which stimulates discussion and builds argumentation skills” [1].

Researchers also pay considerable attention to activating creative activity and increasing motivation. Morozov runs project‑based lessons in which students act as scriptwriters, directors and designers, using neural networks to create historical comics, write dialogues between historical figures, and develop scripts and visual sequences for short videos [7]. Masimzade actively integrates project‑based learning into the digital environment. In the project “Time travel: Russia and the world in the 16th‑17th centuries”, students create interactive maps, timelines, posters, presentations and digital posters. “Analysis of academic performance showed that the introduction of project‑based activities increased student engagement by 50 %” [5]. Mishenina suggests using the “Akinator” service, which guesses a character: “In history lessons, it is useful to use this service to review the activities of various historical figures. An additional motivation for students is the situation when they guess the historical figure faster than the neural network” [6]. Koroleva also notes that AI can stimulate discussion by offering scenarios or questions for debates. “For example, chatbots can act as moderators, generating arguments and counter‑arguments on given topics. This helps students develop argumentation and critical thinking skills” [4].

Alongside optimistic assessments, researchers point to serious risks. Zabelina, Pinchuk and Gritskova highlight the following ethical issues: the protection of personal data and confidentiality; algorithmic bias (AI learns from existing data and may reproduce the stereotypes embedded in them); teachers’ low level of awareness; insufficient material and technical resources; negative attitudes among teachers and parents; and a lack of personal interaction between teacher and student [8]. A particularly critical stance is taken by Gerasimov, a historian. Based on a comparative test of ChatGPT 3.5 from OpenAI and the domestic GigaChat from Sberbank, he concludes that both AI systems “give a mostly negative assessment of Russian history” [2]. He notes that GigaChat makes factual errors (for example, claiming that Ivan the Terrible established the patriarchate in Russia) and often employs liberal and anti‑Soviet assessments. The author concludes: “I consider it inappropriate to recommend either AI for use by schoolchildren and students as an auxiliary tool for studying national history” [2]. At the same time, Fitzner emphasises that “AI is a tool designed to help the teacher, not to replace him or her. The teacher plays a decisive role in interpreting the information obtained with the help of AI, stimulating discussions and summing up results” [1]. Koroleva also warns: “No artificial intelligence guarantees 100 % quality of the output it produces, and neural networks are a supplement to other teaching aids” [4].

Synthesising the data presented in the sources suggests a positive dynamic when AI is used systematically in history lessons. Masimzade reports the following results of her methodological system: the average grade in history and social studies rose by 12 % over two years; the number of students choosing history and social studies for the Unified State Exam increased by 20 %; students’ digital literacy improved by 40 %; critical‑thinking levels rose by 35 % according to standardised tests; and the time spent on educational activities outside of lessons increased by 50 % [5]. Klimentyev, although not providing the final quantitative data from the control stage in the excerpt presented, notes “a positive dynamic in the development of students’ critical thinking, expressed in an increase in the number of students with a high level at the final stage and a decrease in the proportion of students with a low level of critical thinking” [3]. Morozov sums up his experience: “The use of artificial intelligence technology, the differentiation of tasks and the visualisation of learning in history lessons helps to activate the educational and cognitive activity of students and, as a result, fosters the formation of knowledge, skills and abilities” [7].

Based on the analysis of sources, the following methodological recommendations can be made. It is necessary to follow a visualisation algorithm: define the objectives, select the terms, choose the mode of presentation, create the visual materials, and carry out mandatory reflection [7]. Various techniques should be used: “reviving archives”, “dialogue with the era”, “creating historical hypotheses”, “taxonomy of questions” [7]. It is important to foster a critical attitude towards the outputs of AI: students should analyse and verify the texts generated by neural networks, identifying factual errors and bias. AI should be combined with traditional methods, remembering that neural networks are a supplement, not a replacement for textbooks, the teacher’s explanations and live communication. Finally, teacher training needs to be improved: as shown by the survey conducted by Zabelina et al., most teachers have a basic or minimal level of AI proficiency and need methodological materials [8].

Artificial intelligence technologies open up new horizons for school history education. They make it possible to create an “immersion” effect in a historical era through audiovisualisation; to implement a personalised and differentiated approach; to develop critical thinking through the analysis and verification of generated texts; and to increase student motivation and creativity through project‑based learning and game elements. However, as Fitzner stresses, “the human factor, the wisdom and the pedagogical talent of the teacher remain the foundation of the educational process” [1]. AI does not replace the teacher but becomes the teacher’s assistant, expanding the available toolkit and freeing up time for more creative and educational tasks. A key challenge remains the ideological bias and factual unreliability of many AI systems, especially when addressing certain historical topics. Consequently, training teachers to use AI effectively and responsibly, and developing students’ critical thinking and media literacy skills, is of particular importance. Only under these conditions will AI become not a source of falsification but an effective tool for understanding the past.

References:

1. Fitzner, Yu.A. (2026). Artificial intelligence in history lessons: new opportunities for education. In Development of Modern Education in the Context of Pedagogical (Educational) Competenciology: Proceedings of the VI All-Russian Scientific Conference with International Participation (Cheboksary, March 20, 2026) (pp. 92–94). Cheboksary: Publishing House «Sreda».
2. Gerasimov, G.I. (2024). History lessons from artificial intelligence. Ideology of the Future, (13), 89–94.
3. Klimentyev, N.D. (2025). Developing students' critical thinking based on the YandexGPT neural network in history teaching. In *Pedagogy and Psychology in the 21st Century: Current State and Research Trends: Proceedings of the XIII All-Russian Scientific and Practical Conference of Students, Master's Students, Postgraduates and Young Teachers* (Kirov, April 25, 2025) (pp. 612–629). Kirov: Interregional Center for Innovative Technologies in Education.
4. Koroleva, T.V. (2024). Methods of using neural networks in history and social studies lessons in the educational process of a college. In Technological Innovations and Scientific Discoveries: Collection of Scientific Papers Based on the Materials of the XVI International Scientific and Practical Conference (Ufa, November 14, 2024) (pp. 251–255). Ufa: Scientific Publishing Center «Vestnik Nauki».
5. Masimzade, N.V. (2024). Digital transformation of history and social studies lessons: integration of ICT, critical thinking and artificial intelligence for the formation of 21st century competencies. In Science, Education, Innovations: Current Issues and Modern Aspects: Collection of Articles of the XXV International Scientific and Practical Conference (Penza, December 10, 2024) (pp. 96–98). Penza: Nauka i Prosveshchenie (IP Gulyaev G.Yu.).
6. Mishenina, M.V. (2023). Using neural networks to increase motivation in history and social studies lessons. Interactive Science, (10(86)), 40–43.
7. Morozov, A.I. (2026). The use of neural networks and artificial intelligence technology in history lessons for the development of students' subject competencies: audiovisualisation, differentiation of tasks, creativity. In Science Research 2026: Collection of Articles of the II International Scientific and Practical Conference (Petrozavodsk, March 09, 2026) (pp. 55–59). Petrozavodsk: International Center for Scientific Partnership «New Science» (IP Ivanovskaya I. I.).
8. Zabelina, S.B., Pinchuk, I.A., & Gritskova, L.S. (2025). Artificial intelligence in education: history and development trends. World of Science, Culture, Education, (6(115)), 229–233.