Editor’s note: “TIRF Perspectives” is a new publications activity from TIRF. It profiles recent research and developments in English Language Teaching (ELT)—in a practical and accessible format. TIRF Perspectives is aimed at parents, teachers, school leaders, and all student advocates who want to review recent educational trends in a non-academic way. Each series of posts has a common theme, and the theme for this post is the impact of Generative AI on ELT, authored by Michael Carrier, TIRF Trustee and Highdale Learning (London, UK).

Everyone is talking about AI – it’s beginning to get boring.

A huge percentage of talks at English Language Teaching (ELT) conferences is about Generative (Gen) AI – how to bring it into the classroom, which tools to use for conversation practice, or when to support writing skill development.

The debate over AI in ELT reminds me of the 90s when we first talked about using the Internet in class. Pioneers believed in the promise of online learning, many teachers felt it was scary, and academics asked, but where is the theory behind it?

Fast forward to Covid-19 when the Internet became tightly integrated into everyone’s teaching and learning. English schools worldwide now offer Zoom-based classes for practical, pedagogical, and cost-saving reasons. And with great success! Video conferencing is ideal for expanding conversation practice opportunities.

So, the conundrum isn’t new – AI is another technology that hovers at the ELT classroom door for a while, then storms in like a gaggle of teenagers after lunch, disrupting the sedate lesson plan the teacher had in mind.

And as with the Internet, we say things like we have to adapt to it, or it’s not going away, or we can’t let the tools dominate the language learning process.

So far, so familiar. But with GenAI and LLMs (Large Language Models, the text databases that facilitate AI processing), something is different. Yes, we can propose the benefits GenAI might deliver for English learners. But there are also ethical dimensions we need to discuss.

Where We Are

Using the Internet for ELT didn’t threaten anyone’s livelihood – but AI could lead to job losses. AI services do some of the work that teachers do, with bots patiently tutoring for hours on end. AI is already replacing office workers, so shouldn’t English instructors worry about their job security, too?

Also, using websites to teach English didn’t negatively impact learner’s study skills – but AI threatens the development of critical thinking. Writing is the language skill most at risk because students can quickly train ChatGPT or Gemini to research and author a thoughtful assignment designed to include English learner errors.

Even wider than teacher job security and student learning issues, AI raises ethical concerns for society at large. AI tools create credible content – photos and illustrations, texts and stories – that impact the livelihood of people in creative fields. Photographers, writers, and voiceover actors whose work is at the heart of ELT lessons worldwide will lose income.

Notes the UK Society of Authors: “The AI development race is opaque, unfettered and unregulated, and driven primarily by the profit motives of large corporations, despite some likely adverse impacts.” They’ve devised a ‘Human-Authored’ logo to be printed on books that authors can certify as human-, not AI-written.

Where We’re Going

Schools and universities are scrambling to develop ethical AI policies. Some schools post a traffic-light graphic of activities students are encouraged to—or prohibited from—doing. Cambridge University Press offers AI Readiness for Schools support. For a summary of ELT guidance, see The International Research Foundation’s (TIRF’s) paper: AI Ethics: Policy Guidance for Language Education. 

But are AI-support guidelines working? Yes and no. Take writing. There’s a movement to go back to handwriting on paper in class, circumventing the use of AI. But this cumbersome solution doesn’t scale. On the horizon is software that can compare handwritten samples with digital writing assignments from the same student to judge authenticity. HyperWrite’s Similarity Checker is one helpful tool teachers in your institution can try to assess the provenance of student work.

So, are ELT instructors sliding around the decks of the AI Titanic with icebergs ahead? No, there’s time for each teacher to develop a proactive approach to deploying AI. Universities will need to adapt the most as their learning model is based on training students to write well. Language classrooms that rely (as they should) on oral interactions won’t be impacted as much.

Over to You

94% of UK undergraduates use GenAI for assessed work. Our challenge as teachers is to help students use AI productively in ways that don’t undermine their learning. This involves teachers giving students strategies for critiquing, adapting and assessing the accuracy and relevance of AI research they gather, and transparently using it as an input to their work – rather than claiming it as their output. TIRF’s 10-Step Guide to AI Ethics gives school leaders suggestions on how to incorporate AI guidance into teacher development.

Institutions will need careful provision of permissions policies – what learners are allowed to do or not and citation policies to show which sources have been used – and how. TIRF’s policy framework can help with that, too.

ELT students naturally ask their teachers for AI-supported speaking practice recommendations. Conversation apps are an area where AI can accelerate oral fluency. Carve out 30 minutes to review the products that complement your teaching style and student’s needs. Perhaps ELSA for targeted pronunciation practice, or Duolingo for an engaging community of learners.

Join the conversation on Substack – click here! How are you deploying AI ethically and productively in your English teaching?

References

Hyperwrite. (2026). Retrieved from: https://www.hyperwriteai.com/aitools/similarity-and-difference-analyzer

Society of Authors. (2026). Retrieved from: https://societyofauthors.org/where-we-stand/artificial-intelligence/

About the Author

Michael Carrier has worked in language education for many years worldwide as a trainer, author, and director. He has held senior positions at the British Council, Eurocentres and International House. Michael has written several books about language and language teaching, including as co-editor of Digital Language Learning and Teaching: Research, Theory, and Practice (TIRF & Routledge).