In a rapidly changing world, technology plays an important role in our lives across various areas, including education. We now witness the profound development and integration of artificial intelligence (AI) in this field.
 

AI redefines possibilities and challenges us to rethink the future of our interactions, particularly in language learning, where its integration revolutionizes education by offering numerous opportunities for both learners and teachers worldwide to enhance the learning and teaching experience. The countless possibilities of AI take language learning a step further into innovation and effectiveness, facilitating language acquisition and fostering a more interactive and engaging approach to mastering communication skills.

Generative AI has been crucial for all these innovations. Teachers have been using it to explore new ways to improve teaching and learning, integrating it not only in class preparation but also in conversation classes and practice. The emergence of dialogue-based computer-assisted language learning (DB-CALL) has introduced a new way to teach and learn in conversation classes.
 

While the development of this powerful tool and its constant integration into language learning has shown positive effects, researchers caution about potential threats regarding the role of the teacher and the language acquisition process that may be affected.

AI and language learning
 

AI technology in language teaching is based on three main principles: Adaptive learning, Speech recognition, and Natural Language processing.

AI-powered technologies are today considered to be beneficial to language learners as they are capable of offering a personalized learning experience for the user. This means that AI algorithms can tailor the individual personal experience by analysing its own needs, strengths and weaknesses (Patino, 2023) (in Grammarly per example) as well as adapting to their learning rhythms and styles in the creation of exercises contributing to a “more effective and efficient language acquisition”  (Way with words, 2024).
 

Speech recognition involves capturing the learner's oral expressions and transcribing them into text, analysing and giving feedback on the oral communication elements like pronunciation, intonation, and rhythm of speech (Zhong, 2024).

Also integrated into AI, Natural language Processing (NLP) opened new and effective ways for language learning, especially in chatbots like GPT-4 that enables it to understand, interpret, and answer to humans, creating a real-time conversation (Patino, 2023) helping learners to have a better knowledge of the meaning and correct usage of the language they are learning (Zhong, 2024).
 

Today, a wide range of language learning apps and chatbots are available for this purpose. Please refer to Annexe 2 for an overview of the most popular options used worldwide.

A significant number of people are now using or familiar with language learning apps. AI created an impact not only on learning but also on the teaching practices of teachers. It has been integrated into language learning through different forms that are familiar to a lot of people, by the way of apps, chatbots, and generative text.
 

The mobile app Babbel conducted a survey where 87% of the users said that their language skills improved by using the app and its personalised lessons (Patino, 2023).  A study conducted by Xu et al. (2023) demonstrated that Chinese English as a Foreign Language learners who received AI-based instruction exhibited significant improvements in their second language speaking skills compared to those who received traditional methods. Regarding the personalised learning paths, another study stated that this feature helped improve the retention rates of 78% of digital language learners questioned (Astute Analytica, 2024).
 

These studies highlight AI's significant impact on language learning. The results indicate a promising trend in language education, where digital tools and personalized approaches play a crucial role in helping learners achieve their linguistic goals. This trend has led to increased investments in AI for education in recent years, with a notable growth of 25% from 2020 to 2021, of which 40% was related to technologies involving personalized learning paths (Astute Analytica, 2024).
 

AI-powered technologies and foreign language learning: What are the capabilities and benefits?
 

Besides the benefits of the personalised learning experience allowed by the adaptive learning principle, the rise of AI in this field is explained by multiple assets that enhance not only the learning but also the teaching experiences.
 

One of the major benefits of AI-powered tools is that they are 24/7 available for the students as they are online, AI tutors, chatbots and apps can be used at any time of the day, allowing learners to study the language outside the classroom with a flexible time and learning pace aspect.
 

Accessibility is also important as some of the tools are free, have free features or are affordable and allow learners to have access to the materials at any moment and anywhere (Dugošija, 2024).
 

One other benefit of AI-powered tools in foreign language learning is the interactive language practice tools, boosted by the NLP. This asset impacts the effectiveness of learning for the student by creating exercises adapted to the level, style and speed of learning, simulating real-life conversion and giving live and immediate feedback correcting and explaining mistakes (Way with words, 2024) (Rebolledo Font de la Vall & González Araya, 2023; Xie, Chu, Hwang & Wang, 2019, cited in (Dugošija, 2024).
 

Learning a foreign language implies developing the skills in reading, listening, speaking and writing in the language we are learning. Through the different principles of AI-powered technologies for language learning, apps and chatbots are capable of approaching these four crucial communication skills in foreign language learning (OCDE, 2024). The different tools that exist today are useful to work on one or more than one of these language competencies through different types of ways, from game-based approaches to grammar and online dictionaries or even text or dialogue generators (See annexe 2).
 

As said before, Generative AI brought a new dimension to language learning, new but rapidly evolving, this resource is opening new ways to enhance language learning. Through the Generative AI, notably, DB-CALL, teachers and learners address identified challenges, regarding learning a second language, in particular the complexities of speaking, limited opportunities for practice with native speakers, and the lack of feedback in formal education (Goh & Burns, 2012; Divekar et al., 2021, cited in Ericsson & Johansson, 2024).
 

Tutors to facilitate learning – AI conversational agents
 

The recent development of GenAi is particularly being used in conversational classes as a language partner. These AI conversational agents also called, intelligent tutoring systems or virtual assistants take into account the difficulties of learning a second language by creating a virtual learning environment.
 

Here, learners can practice the language they are learning by interacting with the AI conversational agents powered by AI and NLP either through text or speech (Bibauw et al., 2022; McTear, 2021, cited in Ericsson & Johansson, 2024).

Conversational agents, in the form of avatars, serve as language tutors. They are not only able to enhance the learners’ speaking skills but also their willingness to learn, boosting their confidence.  Students feel more comfortable chatting in English with the bot than with their teacher or classmate (OCDE, 2024, cited in Yang et al 2022).
 

These systems analyse each learner's unique strengths and weaknesses, crafting tailored learning paths that address individual needs. As students engage in conversations, the AI provides real-time feedback on pronunciation, grammar, and vocabulary usage (through automated speech recognition) This immediate guidance allows learners to make instant corrections and improvements, accelerating their language acquisition process. By adapting to each user's proficiency level and learning style, these AI companions create an optimsed, efficient, and engaging language learning journey (Meniado, 2023) (OCDE, 2024).
 

Teachers can choose between two types of Chatbots, virtual scripted voice systems, that offer oriented specific tasks for a more structural learning or voice systems based on AI like Siri; which are less task-orientated and structured but encourage language development more naturally, offering less structured learning opportunity rich in authentic interaction (Godwin, 2022).
 

In the region of Nouvelle Aquitaine in France, have been using conversational chatbots like Captain Kelly in elementary school, with whom the students have been practising English in conversational classes, training their oral comprehension and pronunciation. A proof that this can be used and adapted to every level and age.

AI virtual assistants for teachers

In the other way, these AI systems can be used as support or assistants for teachers and trainers in administrative issues and class planning.
 

Administrative work is a big time-consuming activity for teachers. AI apps and chatbots exist today to ensure an important helping hand in administrative support. This support is done through data analyses on the performance of the learners to identify trends and adjust teaching strategies or even facilitating the communication between the students, teachers and parents if the case, by, i.e creating alert notifications on tasks or updates in the students’ progress (Verge AI, 2024).

Virtual assistants for teachers are capable of planning and creating and plan lessons. Through a study made by Meniado (2023), it is shown that ChatGPT can support English teachers in this area by planning lessons for a specific group of learners, regarding their type and level of education. Lesson planning using chat GPT or Quizwizard can assist the teacher in the creation of activities that are adapted to the learning needs styles and levels of the learners.
 

AI tools can generate new and creative lesson ideas for a variety of topics, as well as generate new interactive materials and exercises that can better suit the learners  (Meniado, 2023). This can be done through creating assessments, the generation of questions and tasks, providing feedback and evaluating predefined criteria by the teacher. Again, Quizwizard has been used for this matter since essentially a teacher assistant, helping him to save time by creating exercises from scratch or from support given by the teacher.
 

Innovation in education is considered to be challenging for some teachers however the use of AI combined with their knowledge allows them to plan new lesson ideas that can be modelled to the students' levels (Bonner, Lege, & Frazier (2023) and to specific learning objectives.
 

Please refer to Annexe 3 to see a list of apps and chatbots that can support teachers and learners in conversational classes.
 

The limits of AI in foreign language learning
 

The rise of AI in this field of education brought a lot of advantages but it is also true that with that some limits were identified as well.
 

The use of these AI tools can create an overreliance and dependence on technology because most of the tools do not involve face-to-face human interactions, creating a bubble distant from reality as students can neglect human interaction practices.
 

AI in learning is well-known for its availability compared to a human person; however, it does not replace a teacher. The question remains of how teachers can use AI in their classrooms and what is the role they play in this dynamic. 
 

Furthermore, AI involves the collection of vast personal data, which can raise problems regarding the use of the user’s data, as most of the tools lack regulation on this issue, teachers and learners need to ensure data protection (Zucchet, 2023).
 

One other aspect is that AI, even though they are machines,  is not perfect and cannot have a perfect understanding of cultures and context nuances, as a result, AI-powered tools can make bad interpretations and provide wrong answers in some context-specific situations (cited in, Patty, 2024).
 

Finally, AI also reflects the bias present in their database as this can lead to misrepresentations of languages and cultures and the production of inaccurate content that can reinforce stereotypes (Zucchet, 2023).

Conclusion
 

On a final note, we can talk about this new era of education with the integration of AI in this field, something that is causing a transformative turn in the way we teach and learn.
 

AI offers many new possibilities in foreign language learning inside the classroom, serving as an assistant for the teacher that helps to be more time-efficient and the students to have their own personalised learning experience.

It is important also to mention that teachers and learners should be aware of the limitations of AI in language education as it can have a big and negative impact on our lives in terms of personal data collection, the fact that it can easily be considered as perfect or that it can replace human interaction.
 

Ultimately, AI-powered tools are something that can be beneficial if it’s used by the teacher and the students in a collaborative way, where the human being is still an important figure in this relation. The future holds other surprises, AI continues to evolve offering teachers and learners new ways to enhance language acquisition​