Debates about the influence of artificial intelligence (AI) on all aspects of human life have intensified, particularly with advancements in conversational AI from late 2022 onwards. These rapidly developing technologies offer immediate – and longer term – opportunities, challenges, and threats to educational systems (UNESCO, 2023). However, when it comes to integrating artificial intelligence into education, the current research and guidelines are limited and there is a pressing need for more comprehensive investigation into these matters. This publication was commissioned by the British Council to contribute to emerging research specifically around the impact of AI on English language teaching in education systems.
This report examines three sources of evidence. First, a systematic review of the last 10 years of research on AI in ELT, taking us into early 2024. It then looks to the field, capturing the views of people impacted by these AI technologies. In the global teacher survey, 1,348 teachers from 118 countries and territories reveal how they are using AI and how they feel about the changes it’s bringing. Finally, 19 key witnesses – stakeholders from South and East Asia, Sub-Saharan Africa, the EU and the UK – tell us their views. Bringing together all these sources, the report answers two key questions:
This report details several examples of AI use to promote the four language skills. For instance, in one study, the personal voice assistant Alexa provided a conversational partner for learners to practise speaking (Dizon and Tang, 2020). Automatic speech recognition tools were used to improve pronunciation (Kazu and Kuvvetli, 2023). Grammarly – an AI-powered grammar checking tool – was used to give learners feedback on their writing (Dizon and Gayed, 2021; Nazari et al, 2021). Machine translation tools like Google Translate were used as reference sources for learning to write in English (Chon et al., 2021). In reading, learners improved vocabulary while playing an AI-powered game (Zheng et al., 2015). AI-powered Chatbots helped promote listening skills, though this is a nascent area in the research (Annamalai et al., 2023; Hew, 2023). The majority of teachers in our survey also recognised the ability of AI to support learners’ development in these four skill areas, reinforcing the conclusions drawn from the existing research literature.
Some of the most interesting applications of AI technologies are in the areas of self-regulation and pedagogy. For instance, there is evidence that automatic speech recognition technology could help reduce learners’ anxiety about speaking in public. There are also examples of AI being used to provide a more personalised learning approach using adaptive learning. However, many conventional forms of pedagogy persist within the AI-powered tools design – for example lectures and explanations – and therefore the question remains whether pedagogy is evolving with technology. On the positive side, the research update finds that there is now a much stronger focus on ethics and equity within pedagogy. For detail about affordances in all these areas, see ‘Part I: Background: What the literature says.’
English language teachers report already using AI-powered tools for a range of tasks (76 per cent of teachers from our global survey). Language learning apps, language generation AI and chatbots are the most commonly used tools. Over half of the teachers in the survey reported they are using AI to create lesson material and to help learners practise English. A significant proportion of teachers say they use AI-powered tools to create lesson plans (43 per cent). You can find a detailed breakdown of these results in Part II: The survey: What teachers say.
There are also significant findings relating to which countries are investing in research. Asia currently accounts for two-thirds of the published research on AI in ELT and the number of relevant research articles coming from China increased steeply in 2023 (38 per cent of the articles captured in our review in 2023). In comparison, the US – although widely regarded as leading the AI superpowers – accounted for only 9.4 per cent in the same period.
Perhaps most importantly, the report reveals a concerning skills gap. As mentioned earlier, most teachers report using AI-powered tools (76 per cent), and yet very few report they have had enough training on incorporating it into teaching (a mere 20 per cent). Moreover, AI seems to be impacting language education more than other disciplines, indicating that greater focus is needed on teacher education and training in this field. Stakeholders we interviewed also felt that addressing Al literacy will be a considerable challenge.
A key issue that emerges from our research is inclusion. First, there is the finding that AI can carry messages about what is considered appropriate language use, thereby reinforcing more standardised language use and excluding certain groups or varieties of the language. When it comes to a global language like English, the implication is that practitioners should carefully consider the choice of AI model to ensure particular types of English are not excluded. Then, there’s the potential for AI to worsen digital divides. For example, if AI is widely adopted in better-resourced education systems but not in lower resourced systems, this could widen existing divisions. However, we should also note that AI-powered tools could aid accessibility for some learners, for instance, allowing individuals with visual impairments to use speech to interface with computers.
Yet another issue of significant concern is the ethics surrounding the use of AI, and inherent within AI tools themselves. Our key witnesses emphasised that there is clear evidence of bias and that this needs to be addressed. Some interviewees also expressed concern about Big Tech i.e. the potential influence wielded by large, mostly US-based tech companies on ELT classrooms. It’s clear that there is a need for ethics frameworks and regulation.
AI is bringing other challenges to ELT, such as technology breakdowns (including wrong answers given by the AI), limited capabilities, and learner fear (due to a lack of clarity on how personal data will be used, or of losing the natural learning environment). Cognitive overload and academic misconduct (plagiarism, cheating) have emerged more recently as areas of concern. Clear ethics statements and systems to ensure data privacy may help increase trust and further the responsible use of AI in ELT. It’s important that practitioners remain mindful of limited AI capabilities and are careful about the hype. Overall however, the challenges of AI use in ELT are underreported – perhaps because of the tendency for research to focus on positive findings rather than problems – and more focus is needed on uncovering and mitigating these risks.
When it comes to AI replacing humans, the majority view is that AI will not replace the need for human teachers any time soon. Most teachers in our survey also disagreed when asked whether by 2035, AI could teach English without a teacher, and whether automated translation would make learning languages unnecessary. But many teachers also remained neutral, indicating some uncertainty about the enduring value of language learning and the teacher’s role.
Finally, the question of assessment presents both an opportunity and a threat. Concerns around preventing cheating with AI could lead to the creation of new and innovative assessment tasks. However, the report also raises concerns about the ‘datafication’ of education, and the potential for AI-powered tools to enable the collection of new types of personal data i.e. emotion AI.
This report provides a much-needed overview of the field, including a first-of-its-kind formal study that examines AI and ELT across all learner levels. Artificial intelligence is already impacting English language teaching, learning and assessment; this is clear from the recent rapid advances in AI technologies and their increasing availability. Educators worldwide are grappling with questions about AI use in education more broadly – about relevance, ethics, and access, amongst others. Undeniably, AI will continue to shape our field in the foreseeable future. This report outlines the major considerations and forward-looking measures needed to ensure that AI adoption in ELT moves in the right direction for all stakeholders involved.