The narrative around generative artificial intelligence (GenAI) in education is shifting. We have moved past the initial era of bans and restrictions and are now entering a phase of rapid, widespread adoption. While challenges remain, the current goal is to find a balanced way to bring these powerful tools into the everyday practices of English language teachers and learners. If 2024 and 2025 were the years of exploration and pilot projects using GenAI in the classroom, 2026 will be the year of a more guided integration as institutions consolidate frameworks and strengthen AI literacy.
Conditions
The first steps to be able to do this are to become aware of AI’s functioning and related issues, and to identify what AI is really good at in general – for English language teachers and learners in particular.
We believe GenAI’s inclusion in education cannot be stopped. It is essential to focus on a conscientious, critical, ethical and responsible approach, one that takes into consideration the issues associated with AI and builds on strategies to minimise the negative aspects to unlock the positive ones. In practice, at a macro level, this implies establishing clear institutional and classroom guidelines, providing teacher training. At the same time, we teach students how AI systems generate outputs and how to analyse them critically.
Despite its potential, AI integration faces several critical hurdles. Because models are trained on human-generated data, they often mirror and amplify societal biases, creating ‘black box’ systems where decision-making is opaque and difficult to audit. Furthermore, GenAI’s reliability is limited by its data-driven nature; it often struggles with nuance and creativity, leading to ‘hallucinations’ or illogical outputs. Beyond technical flaws, the environmental footprint of AI, including massive energy and water consumption, e-waste and resource extraction, poses a significant sustainability challenge. In the context of English language teaching, there is a risk of linguistic homogenisation, where diverse varieties of English are sidelined in favour of standardised norms. Finally, the rapid evolution of technology has created a regulatory vacuum, raising concerns over data privacy, intellectual property and an intensifying ‘digital divide’ that leaves under-resourced communities behind. It is important to give visibility to this in our classroom practice and to discuss it both with colleagues and students. We need to understand that AI outputs are not neutral and that analysing them helps students to develop critical awareness; teachers act as mediators who question AI outputs rather than delegate authority to GenAI.
We believe that when certain conditions are met, GenAI can prove helpful for English language teachers and learners. To reduce the downsides, we propose that AI literacy tasks must be at the heart of any GenAI introduction and integration – tasks that foster critical thinking and analysis of AI outputs to raise awareness of issues. Another layer of AI integration is modelling its use and bringing codes of conduct into classrooms. Yet another condition that must be met is the human-in-the-loop (HITL) principle, which means humans actively collaborate with AI models, providing essential feedback, judgment and oversight. This improves accuracy, handles complex edge cases and ensures ethical, reliable outcomes. We must blend human nuance with machine efficiency for better decisions.
The four essentials for AI integration in ELT
Moving from exploration to sustained classroom use requires more than individual experimentation; it calls for shared pedagogical reference points. To support teachers and institutions in making informed decisions about AI use, we outline below the four conditions signalled above that help ensure generative AI enhances rather than undermines teaching and learning in ELT contexts.
AI literacy for teachers and learners
- Understanding limitations, hallucinations, training data
- Distinguishing support vs substitution
Human-in-the-loop pedagogy
- Teacher as final decision-maker
- Students required to justify or critique AI outputs
Task design that demands thinking, not output
- Comparison, reflection, adaptation tasks
- Process over product
Clear classroom norms and codes of conduct
- Transparency (Was AI used? How?)
- Boundaries (what is allowed, what is not)
Teachers’ benefits
Optimising the educator’s workflow
One of the most significant advantages of AI integration is its capacity to serve as a ‘force multiplier’ for teachers. By automating the more labour-intensive administrative aspects of the profession (such as: generating customised lesson plans; curating levelled reading materials; drafting rubrics; creating vocabulary exercises tailored to specific student interests; managing attendance data and maintaining meticulous student records), AI effectively returns the gift of time to the educator. Instead of being buried under a mountain of paperwork, teachers can reallocate their energy toward high-impact activities like one-on-one mentoring, social-emotional support and facilitating deep-dive discussions that require human empathy and cultural nuance that AI cannot handle. None of this, however, can be achieved without proper training in GenAI basics and prompting, as well as a solid teaching knowledge background that can guarantee an informed critical analysis of GenAI’s outputs and suggestions. Using GenAI as a pedagogical design partner can significantly boost a teacher’s potential, while foregrounding their expertise, agency and judgement.
Elevating the cycle of assessment
Furthermore, AI is revolutionising the feedback loop, which is critical in language acquisition. Traditional grading is often delayed, but AI-driven tools could provide instantaneous, granular feedback on grammar, syntax and pronunciation, provided the AI model’s training objective matches the students’ goals, background, age and language level. This immediate reinforcement can help learners correct errors in real time before they become ingrained habits. For the teacher, this means the ‘heavy lifting’ of objective grading can be handled automatically, allowing them to provide more sophisticated, qualitative critiques on a student’s voice, rhetorical style and argumentative depth. And GenAI can identify specific linguistic patterns or recurring struggles that might escape a human observer in a large class. Again, we believe the feedback process cannot be fully automated and the teacher’s participation and decision-making in the process is essential.
Enhancing teacher reflection and professional growth
AI tools can also support teachers’ ongoing professional development by acting as a reflective aid rather than an authoritative source. For example, teachers may use AI to analyse patterns in student errors, generate alternative explanations for challenging language points or explore multiple ways of framing a task or concept. This can prompt reflection on instructional choices and encourage pedagogical experimentation. When approached critically, such interactions position AI as a catalyst for teacher learning, helping educators refine their practice, articulate their pedagogical reasoning and remain responsive to evolving classroom dynamics. This reflective use of AI aligns with a lifelong learning perspective for teachers, positioning them as active learners who continuously question, adapt and evolve their practice.
Learners’ benefits
Democratising access to quality education
AI could serve as a powerful equaliser in the global landscape of English language learning. For many learners, the primary barriers to fluency have historically been a lack of high-quality instruction, insufficient exposure to the target language and few opportunities to practise outside the classroom. AI tools dismantle these obstacles by providing a ‘24/7 classroom’ environment. Whether a student is in a remote village or a bustling city, AI ensures that the lack of a local human expert does not dictate the ceiling of their potential. We are not advocating for the replacement of teachers, whom we still consider a better option when available.
The era of personalised learning pathways
The dream of a ‘tutor for every student’ has transitioned from theory to reality through AI-powered adaptive platforms and intelligent tutoring systems (ITS). These systems analyse a student’s performance in real time to create a personalised educational journey that adapts dynamically and provides scaffolding practice when necessary. AI enables the customisation of curricular resources to meet local needs, ensuring that the content is culturally relevant and linguistically appropriate for specific demographics. This moves education away from one-size-fits-all and towards a model of individual mastery.
Inclusion and Universal Design
AI-driven accessibility tools are transforming the experience for learners with diverse needs. Voice assistants, real-time translation tools and adaptive content-delivery systems ensure that learners with disabilities or those in under-resourced, remote areas are not left behind. For multilingual learners, AI can bridge the gap between their home language and English, providing a safety net that encourages exploration rather than withdrawal.
Targeted skill development
AI can provide specialised support for the specific mechanics of the English language.
- Speaking and pronunciation One of the greatest hurdles for English learners is ‘foreign language anxiety’. AI conversational partners provide a judgement-free zone where students can practise speaking without fear of embarrassment. Automatic speech analysis from appropriate models offers immediate, objective feedback on phonemes and intonation, allowing students to refine their accent and fluency in private before taking those skills into the real world.
- Writing Beyond simple spell checkers, modern AI writing assistants help students move from basic competency to stylistic nuance. By offering suggestions for lexical variation, grammatical accuracy and tone, these tools teach students how to write rather than just correcting their mistakes. They handle the ‘mechanical’ feedback, allowing the student to focus on the clarity of their ideas.
- Reading and vocabulary acquisition AI enhances reading comprehension by embedding vocabulary support directly within the text. Through game-based environments and context-aware definitions, AI helps students acquire language naturally, ensuring they remain in a state of ‘comprehensible input’ where the material is challenging but never overwhelming.
Cultivating learner agency and self-regulation
Perhaps the most overlooked benefit of AI is its role in fostering student autonomy. By providing clear data on their own progress, AI tools encourage students to take ownership of their learning. Students learn to set realistic goals, track their own milestones and manage their study time effectively. This development of self-regulation skills can transform the student from a passive recipient of information into an independent, lifelong learner.
Conclusion
The integration of AI into the English language classroom marks the beginning of a new era, one defined not by the replacement of the teacher, but by the empowerment of the entire educational ecosystem. While we must remain clear-eyed about the challenges of bias, data privacy and the digital divide, the potential for positive transformation is undeniable.
For the teacher, AI acts as a sophisticated digital partner, absorbing the weight of administrative logistics and objective grading. This shift does more than just reduce burnout; it restores the educator to their most vital role: a mentor, a cultural guide and an orchestrator of meaningful human connection. When the ‘mechanics’ of teaching are streamlined, the ‘art’ of teaching can truly flourish.
For the learner, AI provides a personalised, judgement-free sanctuary where the fear of making a mistake is replaced by the curiosity to explore. By democratising access to feedback and tailoring the curriculum to individual needs, AI ensures that the path to English proficiency is no longer a one-size-fits-all struggle, but a dynamic journey toward self-regulation and independence.
Ultimately, the most successful AI-driven classrooms will be those that use technology to amplify human potential. As we move forward, the goal is clear: to leverage these powerful tools to build a world where language is no longer a barrier but a bridge, supported by the precision of AI and brought to life by the empathy and insight of the teachers.
If we consider preparation for the future as an essential goal of education, integrating AI into the curriculum helps equip students with essential AI literacy, computational thinking skills and understanding of AI’s societal impact, preparing them for future academic and career pursuits.
As we stand at this digital crossroads, the question is no longer if AI will change English education, but how we will choose to use it to make our classrooms more human than ever before.
