The use of generative artificial intelligence in the study and teaching of foreign languages: the results of a systematic analysis

Gruzdev Andrey Vladimirovich Lecturer; Department of Foreign Languages; Moscow Institute of International Relations (Odintsovo branch) Novosportivnaya str., 3, Moscow region, 143007, Russia a.gruzdev@odin.mgimo.ru

Introduction

Artificial intelligence (AI) is a field of computer science that deals with the creation of intelligent computer systems capable of performing tasks that have traditionally been considered the prerogative of the human mind ^[12]. In the last decade, AI technologies have made significant progress thanks to the development of machine learning methods, big data, cloud computing, and the development of generative models. These achievements are widely used in various fields, including medicine, finance, transport and, of course, education ^[2].

One of the most promising areas of AI use is the study and teaching of foreign languages. Modern AI systems are able to recognize speech, translate texts, assess language proficiency, and even simulate conversations with users ^{[2, 5, 12]}. This opens up new opportunities for improving the educational process, making it more efficient, accessible and personalized.

The dramatic progress in the use of AI in the study and teaching of foreign languages is associated with the emergence and rapid spread of generative models. Generative AI is the use of AI to create new content such as text, images, music, audio, and video, applying a machine learning model to explore patterns and relationships in a dataset of human-created content, and then using learned patterns to generate new content. Unlike previous forms of AI technologies that use machine learning algorithms and prediction based on past behavior, generative AI focuses on creating new textual and multimodal content using large language models (LLM), models that create images, audio and video ^[12].

Since the launch of ChatGPT by OpenAI in November 2022, generative AI technologies have attracted worldwide attention and demonstrated the potential to change the landscape of education, including language. In the traditional academic environment, language teachers acted as the main facilitators of language acquisition and language skills development, and students mastered language skills under the guidance of teachers. However, the advent of the Internet, search engines, and now generative artificial intelligence programs marks a new stage in the transformation of language education, often reducing the role of teachers ^{[6, 7]}.

The purpose of this article is to review the current state and prospects of using generative AI in learning and teaching foreign languages based on a systematic review of relevant scientific literature. We will analyze existing approaches and methods, as well as discuss their advantages and limitations. Finally, we will look at the challenges and directions of future research in this area.

Although there has been a significant body of published research and literature reviews on artificial intelligence (AI) in recent years, the existing literature specifically devoted to the application of generative AI in language education, as well as generative AI in general, is still very limited. Given that generative AI is a relatively new and rapidly developing technology, a review of the current state of research is extremely necessary.

Research methodology

The empirical material for this article is scientific and popular science texts focused on the integration of generative AI into language education. The criteria for selecting sources included the year of publication (from 2017 to 2024) and the type of publication (theoretical articles, empirical research, review articles, technology reviews, editorial opinions, and discussion articles). At the same time, publications without a focus on language education or generative AI, preprints of already published papers, student assignments, commercial websites, blogs, journal articles, conference abstracts were excluded.

Although the search covered the period from 2017 to July 2024, only one entry was published in 2022, and all the others in 2023 and 2024, which is explained by the fact that generative AI in language education began to attract attention only after the launch of ChatGPT in November 2022. A total of 44 articles were selected for analysis.

By geography of origin, the selected studies include Asian countries, Middle Eastern countries, European countries, and Russia.

The main conclusions

The landscape of research on the use of generative AI for teaching and learning foreign languages

In the literature, generative AI is usually described as an advanced technology based on large language models (LLM), existing in the form of natural language processing (NLP) software, for example, chatbots. In the context of language education, this technology falls into four categories:

— AI-based learning tools;

— chatbot-based learning tools;

— AI-based dialogue systems for learning a foreign language;

— AI-based platforms and applications ^{[1, 3, 9, 10]}.

Most of the publications focus on learning English as a Foreign Language (EFL) ^[6]. The use of generative AI is being explored at various educational levels:

— pre-school education;

— primary school education;

— secondary school and secondary vocational education;

— higher education;

— additional education and, in particular, preparation for international language tests.

For language teaching, there are several main areas of research into the use of AI:

— General teaching and learning issues;

— the policy of educational organizations regarding AI;

— using AI to evaluate work;

— teaching writing in a foreign language using generative models;

— ethical issues of using AI for learning and teaching foreign languages;

— AI errors ^[13].

The most widely studied application of generative AI has been its use for teaching writing ^[8]. Research demonstrates that generative models can help students improve writing skills by providing feedback on grammar, vocabulary, and sentence structure, as well as expand vocabulary and improve sentence structure.

Advantages of using AI in teaching foreign languages

AI has a number of unique features that make it a valuable tool for teachers and students. Let's look at the main advantages of using AI in the process of learning and teaching foreign languages.

Personalizing learning

One of the key features of AI is the ability to adapt to the individual needs of each student. AI systems can analyze learning data such as test scores, app activity, and interaction with learning materials to offer personalized learning recommendations ^[11]. For example, the system can automatically select exercises that match the student's level of training, offer additional materials for in-depth study of certain topics, or focus on weaknesses.

Personalization of foreign language teaching using artificial intelligence is carried out through an integrated approach, including the collection and analysis of data about each student, the use of machine learning algorithms and the creation of dynamic recommendations. Below are the key steps and mechanisms underlying personalization.

The first step in personalization is to collect information about the student. These include:

— language proficiency level: initial assessment, tests of vocabulary, grammar, phonetics and other indicators;

— Preferred topics and interests: analyzing the student's hobbies in order to select thematically relevant materials;

— learning styles: information about whether the student prefers visual, auditory, or practical learning of the material;

— frequency and intensity of training: how much time is a student willing to spend on classes, when and where he prefers to study;

— the results of the assignments: how well the student copes with the assignments, what mistakes he makes, where difficulties arise.

All this data is collected using the platform's built-in tools, applications, or services that use AI.

Based on the information collected, machine learning algorithms are used to analyze student behavior. Algorithms identify patterns, strengths and weaknesses of the student, his pace of mastering the material and preferences. This allows you to determine the individual needs of each student.

After analyzing the data, the AI generates an individual curriculum that meets the specific goals and needs of the student. This includes selecting the optimal level of difficulty of exercises, identifying priority areas for study (for example, an emphasis on grammar or conversational practice), forming a sequence of lessons appropriate to the student's learning style, and recommendations for additional materials (videos, articles, audio) depending on interests and goals ^[7].

The AI continuously monitors the student's progress and updates recommendations in real time. If a student shows rapid progress in one area, the program will offer more complicated tasks. If difficulties arise, she will reduce the complexity or suggest alternative ways to explain the material.

Some students prefer independent study, while others need regular checks and feedback. AI can adapt the program to different learning styles.

Many modern platforms use AI to create interactive elements, such as: chatbots that help improve conversational skills by simulating real dialogues; game mechanics that stimulate active participation in the learning process and increase motivation; virtual mentors who give advice and recommendations during the lesson ^[7].

Finally, the AI generates a detailed report on the student's progress, showing which areas have already been mastered and which still need to be worked on. This report is available to both the student and the teacher, if any. Thus, each participant in the educational process gets a clear idea of the successes achieved and possible directions for further development.

Automation of knowledge assessment

Traditional methods of assessing the level of language proficiency often require significant time from the teacher. AI makes it possible to automate this process by offering instant feedback to students ^[13]. For example, speech recognition systems can analyze students' pronunciation and grammar, providing detailed reports on mistakes made and recommendations for improvement. This helps to reduce the burden on teachers and increase the effectiveness of the educational process.

An analysis of the selected texts shows that generative AI plays an important role in automating knowledge assessment through a variety of tools and technologies that help educational institutions and teachers evaluate students' academic performance faster and more accurately. Next, we present several aspects where generative AI is used especially actively.

Automated test verification — AI is able to quickly check various types of test tasks: from questions with a choice of answers to essays. This significantly saves teachers time and allows them to focus on more complex aspects of the educational process. Answer selection tests are automatically checked using algorithms for matching correct answers. Natural language processing (NLP) methods are used to check essays and open-ended questions, which evaluate the semantic content of the text, identify key ideas, and even check grammar.

Personalized Recommendations and Adaptive Learning — AI helps to create personalized curricula and recommendations for each student based on an analysis of previous results and student behavior. Such systems allow you to better track progress and offer assignments that match the level of training of a particular student. For example, if a student is having difficulty with a particular topic, the system will offer additional materials or exercises specifically on that topic. Educational platforms accumulate huge amounts of data on student performance, including tests, homework, class attendance, and other parameters. AI processes this data, helping to identify patterns and trends in academic performance, which contributes to the development of more effective educational strategies.

Predicting exam results — AI algorithms are able to predict the results of future exams based on students' past achievements. This allows you to adjust the learning process in time and help students prepare for important tests.

Real—time Feedback - Some educational applications use AI to provide instant feedback to students while completing assignments. This is especially useful for independent work when the teacher is physically unable to be around.

Interactivity and engagement

Modern AI systems offer a variety of interactive learning formats, such as chatbots, conversation practice simulators, and games. Such tools help to increase students' motivation, as they allow them to learn in a playful way, interacting with virtual characters or completing tasks in exciting scenarios ^[7]. For example, students can practice conversational skills by communicating with a virtual conversationalist who adapts to their level and interests. Further, based on the conducted review, we will present the key ways to use generative AI in order to increase student engagement.

Interactive dialogue training — Generative AI allows you to create virtual interlocutors with whom students can practice speaking. These interlocutors can simulate real conversations by asking questions and responding to students' answers. The interactivity of such trainings increases motivation and interest in learning a language, because students can practice in a natural communication environment ^[5].

Creating authentic materials — Generative AI can generate texts, audio recordings and videos in the language being studied, adapted to the specific needs and interests of students. For example, creating news, stories, or dialogues on current topics. Authentic materials make language learning more interesting and applicable to real life.

Game Elements and Gamification — The use of game elements such as quests, puzzles, and rewards encourages students to actively participate in the learning process. Generative AI can integrate into such games, creating unique scenarios and tasks that match the level and interests of the participants.

Simulation of real—world situations - AI can simulate various life situations, such as visiting a restaurant, traveling, or a business meeting. Students can practice in these situations, developing communication skills and understanding language in context. The realism of such scenarios enhances engagement and motivation to learn.

Social Interaction and Collaboration — AI-based platforms can bring students together in groups for joint projects and discussions. Interacting with fellow students through chatbots or virtual classrooms creates a sense of community and support, which also increases interest in learning a language.

Accessibility and flexibility — thanks to AI, students can study at any time and place convenient for them. They can use mobile apps, web platforms, or smart devices to practice the language. The flexibility and accessibility of learning facilitate regular interaction with the material and increase overall engagement.

White spots in knowledge about generative AI in teaching and learning foreign languages

The review revealed several significant gaps in research on the use of generative AI for learning and teaching foreign languages. First, we note a general lack of empirical research. More research is needed to comprehensively understand the short- and long-term effectiveness and impact of generative AI tools, including both textual and multimodal tools and their specific applications in language education.

Secondly, we note the limited discussion of ethical aspects. Ongoing and regular research is required to examine the ethical considerations and potential limitations of rapidly changing technologies. It is especially important to consider data privacy and security issues, a topic that has received insufficient attention in the existing literature. Another ethical aspect is the search for red lines that define acceptable boundaries for the use of AI by students ^[4].

Further, we see a lack of research focused on specific aspects of language teaching and learning. Future research should focus on specific language skills, such as writing or speaking different languages. Such research is needed, in particular, to develop effective interventions in the educational process.

Another blind spot is the limited attention paid to AI literacy. There is a lack of discussion about AI literacy among language researchers/educators. Teachers' knowledge of data privacy and security within the framework of generative AI literacy is of paramount importance, as teachers must inform students about the potential risks of using their data for learning and replication.

Finally, it is worth noting rare examples of collaboration between students, language teachers, researchers, educational administrators, and model developers to ensure meaningful and responsible integration of generative AI into language education.

Conclusions and practical recommendations

The use of generative AI in language teaching and learning is a promising area of research with the potential to transform language education. This article answers a number of research questions by identifying key terms related to generative AI in language education, the most researched languages and levels of education, areas of research, attitudes towards the use of generative AI, and the potential benefits and challenges of its implementation. Based on the review, a number of recommendations can be formulated.

Educators should consider including generative AI tools in their teaching practice, while remaining vigilant and attentive to potential risks. Continuous professional development is crucial to ensure informed decisions and effective integration of generative AI tools.

Researchers need to continue empirical research on the effectiveness and impact of generative AI tools, ethical considerations, and educational interventions to develop specific language skills and involve stakeholders in responsible integration.

Finally, society as a whole should take into account that the impact of generative AI goes beyond language education, and it is important to consider its integration into other subject areas.