Generative AI in Language Teaching: A Reflective Perspective from 25 Years in Higher Education

5 January 2026

After 25 years of teaching French in higher education, I have seen trends come and go: from photocopied handouts to cassette-based language labs and now online platforms. Generative AI feels different: it is not just another tool; it is a paradigm shift. In my own classes, whether guiding beginners through their first bonjour or helping advanced learners craft nuanced essays on French literature, I see AI as both a powerful ally and a potential disruptor.

Indeed, Gen AI offers many opportunities for language learning and teaching. In particular, it offers personalisation and adaptive learning at a scale we could only dream of a decade ago. Tools such as Duolingo adjust difficulty dynamically, keeping beginners engaged. For advanced learners, writing assistants like Grammarly and conversational agents such as ChatGPT provide feedback on tone and coherence. These tools can accelerate progress, but cultural nuance and idiomatic expression still require human judgement (Almusharraf & Khlaif, 2024).

AI also enhances engagement and inclusivity. I have experimented with chatbots for role-play scenarios in French and noted they reduce anxiety and encourage creativity by giving students safe, low-pressure opportunities to practice speaking and writing (Zhang et al., 2024). Gamified learning is also motivating. AI can generate quizzes or vocabulary challenges, turning routine drills into interactive games. Students with different learning preferences or needs (visual, auditory, kinesthetic) benefit from AI-generated videos, transcripts or visuals. In practice, AI helps teachers create dynamic lessons where every student feels seen, supported and challenged. These developments align with policy recommendations for widening participation and supporting diverse learners (DfE, 2024).

But Gen AI also introduces new challenges, complexities in academic integrity and ethical dilemmas. I have encountered written work in the target language where AI-generated text masked a student’s true proficiency, prompting me to rethink assessment strategies. Akari Software (2025) warns that AI-enabled cheating risks eroding trust and devaluing credentials. Authentic assessments, oral presentation, in-class tests and exams and reflective journals are essential to balance out work completed at home, often with undisclosed AI help.

Bias and cultural sensitivity remain concerns too. Translation tools often misinterpret idiomatic French expressions, which I use as teachable moments. AI-generated texts in the target language can be biased in terms of representation and cultural diversity. Pedagogical redesign is therefore critical: tasks must emphasise process and critical thinking rather than product (Koh & Doroudi, 2023). The Russell Group and QAA advocate for ethical frameworks and transparency, principles that resonate strongly in language education.

There are implications of AI use in language learning and teaching across proficiency levels. For beginners, AI excels at vocabulary building and pronunciation drills, but over-reliance on translation tools can hinder grammar acquisition. Intermediate learners benefit from conversational agents, provided tasks are scaffolded to avoid superficial engagement. Advanced learners can use AI for drafting and editing, but critical evaluation of outputs is vital to maintain originality and voice.

This is why it is crucial to consider pedagogical and policy perspectives. Recent reports stress the importance of Gen AI literacy. Harvard Business School argues that students must learn to use AI responsibly, understanding its limitations and ethical implications. In language learning, this could mean critically reviewing and post-editing AI-generated translations or refining chatbot dialogues to preserve cultural nuance. HEPI and Zenodo call for curriculum reform to embed these skills, while UNESCO emphasises human-centred pedagogy.

But digital literacy also needs to be developed in language educators themselves. AI can generate authentic texts in the target language, simulate conversations and provide feedback, offering powerful opportunities for language practice and cultural exploration. However, without strong digital literacy, language teachers risk either underutilising these tools or failing to guide students in critically evaluating AI-generated content. By investing time in mastering AI, they can enrich pedagogy, foster creativity and personalise learning, while equipping students with critical skills to discern bias, accuracy and ethical implications in digital communication.

It is therefore clear to me that Generative AI complements the human dimension of language learning but it should never replace it. Technology can accelerate progress, but authentic communication, cultural understanding, and critical thinking remain deeply human endeavours. Generative AI can support these goals, but it cannot embody them.

Dr Elsa Petit

Senior Lecturer in French Language Studies

https://www.qmul.ac.uk/sllf/modern-languages-and-cultures/people/academics/profiles/petit.html

References

Almusharraf, N., & Khlaif, Z. N. (2024). Artificial intelligence in language education: A systematic review. Journal of Language Teaching and Research, 15(2), 123–140. https://files.eric.ed.gov/fulltext/EJ1482233.pdf

Akari Software. (2025). The role of AI in academic assessment and cheating. https://akarisoftware.com/2025/03/27/akari-the-role-of-AI-in-academic-assessment-cheating

DfE. (2024). Generative AI in education: Educator and expert views report. https://assets.publishing.service.gov.uk/media/65b8cd41b5cb6e000d8bb74e/DfE_GenAI_in_education_-_Educator_and_expert_views_report.pdf

Harvard Business School. (2023). Are your students ready for AI? https://hbsp.harvard.edu/inspiring-minds/are-your-students-ready-for-ai

Koh, E., & Doroudi, S. (2023). Learning, teaching, and assessment with generative artificial intelligence: Towards a plateau of productivity. Learning: Research and Practice, 9(2), 109–116. https://doi.org/10.1080/23735082.2023.2261131

Pearson. (2025). Ethical challenges of AI in education. https://www.pearson.com/languages/it-it/community/blogs/ethical-challenges-of-ai-in-education-6-25.html

Zhang, Y., Li, H., & Wang, X. (2024). Generative AI as a cognitive co-pilot in English language learning. Education Sciences, 15(6), 686. https://doi.org/10.3390/educsci15060686