Blog: AI in the languages classroom: enhancing, not replacing the teacher’s role
The NCLE recently held a webinar hosted by Joe Dale, NCLE AI Project Lead, showcasing how language teachers are integrating artificial intelligence into their classrooms. In this piece, Joe summarises the key takeaways from the webinar, including insights from three different NCLE Language Networks’ pilot projects demonstrating innovative, practical applications of AI tools for language learning.
Written by Joe Dale
It is important to emphasise that AI is not intended to replace teachers but rather to enhance their capabilities. The focus remains on keeping “the human in the loop” through an 80-20 rule, where AI handles approximately 80% of routine tasks while teachers provide the crucial 20% of professional judgement, critical thinking, and human intervention. This approach allows teachers to check for inaccuracies, bias, and ‘hallucinations’ (errors which AI can sometimes generate), while using their expertise to refine AI-generated content.
For classroom use with pupils, it is also important to stress that we must ensure AI tools are GDPR compliant, age-appropriate, and align with school AI policies. For example, many AI tools require users to be 13 or older with parental permission, so careful consideration of terms and conditions is essential.
Trent and Tame Language Network: Building speaking confidence with Mizou
Chris Parry from the Trent and Tame Language Network presented findings from their pilot project using Mizou, a platform of customisable AI chatbots designed for interaction in the target language. The project addressed two key challenges: pupils’ need for more spontaneous speaking practice as required by the new GCSE specifications; and the common frustration of parents unable to help their children practise at home.
Mizou allows teachers to create custom chatbots tailored to classroom content, with pupils able to interact through speech or text. The platform provides clear pronunciation in the target language and generates transcripts for teacher review.
Chris’s results showed significant improvements in pupil confidence, particularly among quieter pupils who benefited from the non-judgmental environment. One Year 7 pupil demonstrated remarkable progress between sessions, extending responses from brief answers to developed sentences containing multiple verbs. The impact extended beyond speaking skills, as pupils began applying this extended response pattern to writing tasks. Pupil feedback was overwhelmingly positive, with many recognising the learning benefits.
Thames West Language Network: Enhancing assessment analysis with ChatGPT
Emily Lewis from Thames West Language Network described her project using ChatGPT to streamline assessment and feedback processes. Her approach focussed on four key areas: predicting common misconceptions before teaching; creating targeted feedback materials; analysing pupil errors from assessment data; and generating individualised feedback.
Emily demonstrated how ChatGPT could predict potential pupil misconceptions when teaching specific grammar points: particularly valuable for teachers working in languages outside their specialism or for early career teachers perhaps less familiar with some common pupil errors. The AI could analyse assessment mistakes and suggest underlying causes, enabling more targeted remedial work.
The system proved effective at categorising errors from writing assessments and creating differentiated worksheets addressing the most common issues. When generating direct feedback to pupils, ChatGPT provided detailed explanations of errors and suggestions for improvement, offering a level of individualised attention that time constraints might usually prevent teachers from providing themselves.
Kinder Language Network: Increasing motivation through AI-generated music
Catherine Woodward and Michelle Massey from the Kinder Language Network presented their project using Suno, an AI tool that creates custom songs in target languages. Their project aimed to increase motivation and retention of phonics, vocabulary, and grammar through music.
Teachers created songs based on classroom content (such as restaurant problems for Year 10 pupils) then used these as springboards for various activities. Pupils highlighted phonemes within song lyrics and engaged in vocabulary practice through tools like call-and-response exercises. The songs were validated using multilingual profiler tools to ensure 93% alignment with GCSE vocabulary lists, reducing cognitive overload.
Beyond phonics and vocabulary, songs also supported grammar instruction. A Year 9 perfect tense exercise used AI-generated music as a gap-fill activity, helping pupils identify correct past participles. The motivational impact proved significant, with one pupil independently researching the song online and translating the lyrics without prompting, which demonstrated strong intrinsic motivation to engage with the language.
What AI can do for you, what it can’t – and how you can tell the difference
These projects from the NCLE Language Networks show AI’s potential to address several of the most persistent challenges facing languages teachers: providing fruitful speaking practice opportunities outside classroom time; streamlining assessment workload all while improving feedback quality; and increasing pupil motivation through engaging content.
Further, AI tools show promise for supporting differentiation, reducing teacher workload, and providing pupils with personalised learning experiences while preserving the essential human element of language teaching.
However, the webinar emphasised the importance of maintaining a focus on pedagogy over technological novelty. Each application of AI was grounded in teacher-led classroom practice with clear learning objectives, rather than AI implementation for its own sake.
As these pilot projects develop further, they will offer evidence-based models to help guide other teachers who are considering AI integration in their practice in ways that will enhance rather than replace traditional language teaching methods. The findings and resources created by the projects will be freely available to all, hosted on NCLE’s LEO CPD platform.
Explore further with NCLE:
- Joe Dale has authored a specific strand of CPD on AI and EdTech – freely available on the LEO Platform. It is modular, accessible and flexible: easy to explore and complete in the order you want, when you want and where you want.
Findings and resources from the NCLE Language Networks’ pilot AI projects will be published on LEO over the course of the 2025-26 academic year.
Sign up to NCLE’s free CPD platform Language Educators Online (LEO) to gain access to NCLE’s full suite of CPD for languages teachers, including on AI. Registration is free and takes just a couple of minutes.
Joe Dale is an independent modern languages and EdTech consultant based in the UK. He is the AI Project Lead for the NCLE and his new Facebook group ‘Language Teaching with AI’ already has over 6000 members.