The Linguist

Chartered Institute of Linguists SPRING 2026 The Linguist 29 How AI may be helping (or hindering) language students JOANNA BIERNAT Almost all aspects of our lives are now linked to AI. As linguists, we often rely on AI to perform our daily tasks without even realising it. Google utilises AI capabilities such as built- in algorithms, while the social media we use to promote our businesses use AI to organise huge amounts of data and information. Thanks to machine learning (ML) algorithms, email providers instantly identify spam messages, while many of us also use Google Assistant, which uses Natural Language Processing (NLP) to respond to various queries. We often talk about AI in translation and interpreting, but it is also increasingly being used within language learning and research into language acquisition. People seeking to improve their writing, speaking, listening or reading skills in their native or target language often use AI-driven programs. A great example is Grammarly, which offers free "AI- enabled English communication assistance". Language learning apps such as Babbel and Duolingo provide a personalised experience by integrating chatbots. Dr Barbara Konat outlines how built-in probability modules enable chatbots to respond with a degree of compassion. 1 My younger students, in particular, appreciate the impression of being understood. These apps provide individual feedback on language skills and create lessons tailored to users' specific needs. This provides a new area of work for language experts, as communication assistant, e-learning and language learning software companies employ linguists to improve their services. Research into language acquisition draws on linguistics, education, computer science, neuropsychology and other disciplines. Comparing large volumes of data used to be an extremely time-consuming, labour- intensive and expensive process, but AI can do some of this work in seconds. This provides valuable insights in a range of fields, including how languages are acquired across different populations and contexts, 2 and ways to improve the treatment of aphasia. 3 Despite the benefits, there are problems with the increasing use of AI. When working with language learners, I have observed a reduction in creativity as a direct result of their AI usage. The AI-based pronunciation they often turn to is not always correct. Moreover, some perceive AI as a tool that gives them answers to everything; others say learning languages is not beneficial as translation engines and chatbots can do it for them. While researchers use a wide range of language analysis methods, the data used to train AI models is often not diverse enough to deliver a reliable interpretation across different languages and dialects, as Gulordava et al point out. 4 There are also ethical considerations about where the data comes from and how it is being used. More and more of the content on the internet consists of input from artificial language models, leading to the risk that future language models will be generated based almost entirely on AI-created data, rather than by people in their natural language environments. 5 As Sabine Braun notes, AI lacks "understanding of the world, and the social, economic, cultural, political and other factors shaping human language use". 6 AI is strongly integrated within the fields of language research and language acquisition. It empowers language discovery and transforms the way we perceive linguistic research. However, certain issues remain. It is important to challenge AI findings regarding languages, and to check and verify automated analyses. Notes 1 Konat, B (2022) 'Czy sztuczna inteligencja to inteligencja emocjonalna?'; cutt.ly/etbdhpl6 2 Roy, BC et al (2015) 'Predicting the Birth of a Spoken Word.' In Proceedings of the National Academy of Sciences, 112(41), 12663-12668 3 Kuroda, T et al (2025) 'Utility of Artificial Intelligence-based Conversation Voice Analysis for Detecting Cognitive Decline'. In Plos One 4 Gulordava, K et al (2018) 'Colorless Green Recurrent Networks Dream Hierarchically'. In Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human language technologies, 1, 1195-1205 5 Op. cit. Konat 6 Braun, S (2024) The Linguist, 63,3 Dr Joanna Biernat-Sowka MCIL is a linguist, tutor and researcher who works for a multilingual tech company. TL TAUGHT BY THE BOT? © SHUTTERSTOCK

• Cover