The rapid proliferation of artificial intelligence (AI) tools has posed new challenges for academic integrity in higher education. The use of large language models (LLMs), machine translation, and paraphrasing tools complicates the identification of student-authored texts and raises urgent questions regarding authorship attribution in academic foreign language learning. This paper examines the limitations of existing AI text detection tools and presents the results of a pilot experiment involving English as a Foreign Language (EFL) student essays tested against two widely used detectors, ZeroGPT and GPTZero. The study demonstrates that while both detectors achieve high group-level separability (AUC$\approx$0.94), their reliability at the level of individual assignments is unacceptably low, with error rates of up to 29%. These findings suggest that detectors may be useful as supplementary indicators but cannot serve as independent proof of academic dishonesty. The article concludes by arguing for processoriented approaches in assessment, emphasizing documentation of the writing process, oral defenses, and student reflection on AI use.

The study explores a pedagogically driven prompting strategy for generating AI-assisted tasks to enhance mathematical literacy of Russian 5th-grade students. Grounded in international frameworks (NCTM, PISA, TIMSS) and AI-ineducation research, the authors designed a hybrid Few-Shot + Chain-of-Thought approach using ChatGPT-4o. The methodology involved defining metaparameters, generating context-based tasks, and conducting expert evaluation. Results demonstrate that the strategy produces tasks comparable to textbook models, with teachers rating them highly for relevance (M=4.81 / 5) and age appropriateness (M=4.55 / 5). Limitations were observed in AI-generated terminology, indicating accessibility issues for younger learners. The validated approach offers a scalable method for task generation aligned with PISA and national standards, supporting integration into platforms such as the Russian Electronic School and Spherum, and expanding opportunities for teacher-AI collaboration in instructional design.

The article discusses the development of a system that uses artificial intelligence (AI) to generate individualized mathematics assignments for bilingual students in Tatarstan, Russia. The goal is to enhance learning by tailoring assignments to students’ linguistic preferences, cognitive styles, and knowledge levels. The system employs machine learning techniques and GPT-based models to create personalized tasks that align with curriculum goals while addressing linguistic diversity, particularly for Tatar-Russian bilinguals. The study evaluates several large language models (LLMs), including GPT-4, GPT-3.5 Turbo, YandexGPT, and GigaChat, based on their ability to generate math problems and content in the Tatar language. While GPT-4 and GPT-3.5 Turbo show superior performance in producing accurate and semantically correct problems, their proficiency in Tatar remains inconsistent. The research underscores the need for further development of LLMs to enhance content generation for bilingual educational contexts and highlights the potential of AI in advancing adaptive learning for mathematics education. Future directions include expanding the system’s functionality and testing its effectiveness across diverse educational settings.

This article presents an algorithm for assessing text readability based on the analysis of polysemous words. The algorithm was tested on 30 Russian language textbooks for different grade levels, with the total size of the corpus of 1,097,170 words. We estimated values of two complexity predictors, i.e. the number of polysemous words (p1) and number of unique word senses (p2). The research demonstrated a high positive correlation between the respective parameters and text grade levels. The proposed algorithm has a potential to be applied in numerous fields that require text readability assessment and include education, law, medicine and business. The research prospects include validating the algorithm in other languages and text types, as well as contrast it with other text readability assessment algorithms.

The article highlights implementation of Natural Language Processing (NLP) specifically spaCy, to analyze the linguistic complexity and grammatical characteristics of the Russian-English Math CLIL course text. Collaboration between subject and language teachers in planning and delivering CLIL lessons causes the need to develop new methods for selecting and delivering educational content in pre-service teacher training in order to improve pre-service teachers’ English language proficiency and academic writing skills. The methodology section describes the research design, including the use of NLP tools such as spaCy for processing and analyzing text data. The results section presents the findings of the study, including the verb tense ratio, sentence mood ratio, active/passive voice ratio, and sentence type distribution. The article concludes by emphasizing the importance of bilingual and multilingual education, the necessity for bilingual teacher training, and the potential of NLP tools in enhancing language instruction in CLIL contexts.

This article deals with the problem of development of Cyrillic-Latin converter for Tatar language which will be able to convert a text written in Tatar to Latin using Cyrillic graphics. In addition, the article describes some aspects of Cyrillic graphics to Latin for the Tatar language. The authors worked in two directions: various statistical methods necessary for the Converter operation were considered, as well as the speed and accuracy of the conversion algorithms were analyzed.We created an algorithm and developed software modules that allow converting messages written in the Tatar Cyrillic alphabet to the Tatar Latin alphabet.According to the local law acts, a verbal and an algorithmic model of conversion was constructed. In the process of development, it turned out that the process of a Tatar word conversion depends on its origin. Native Tatar words are converted according to the phonetic principle (κeJIaM - qäläm), the borrowed words are converted according to the rules of transliteration. The main problem of the study is the problem of a word origin recognition. In order to solve this problem, the authors propose various algorithmic models. Software tools based on the statistical processing of linguistic data are considered and developed in the work: combined bigram analysis, naive Bayesian classification and a brute-force direct search. Each of these algorithms is used to determine the etymology of a word, on which depends the application of certain rules of conversion from Cyrillic to Latin.Thus, the result of our work is a developed software product that can perform the process of converting Cyrillic graphics into Latin for Tatar language. Further research in this area is related to the development of the software tool and its use in educational activities.