Category: Artificial Intelligence (AI)

Teachers’ use of generative AI: a ‘dirty little secret’?

Barnes, Melissa, and Ekaterina Tour. 2025. “Teachers’ Use of Generative AI: A ‘Dirty Little Secret’?” Language and Education, April, 1–16.

This article reports on a study of how a group of English as an Additional Language (EAL) teachers in Australia perceive and use generative AI in their teaching. The article draws on data from a larger study that examined Generative AI in the Australian education system.

The teachers were asked to respond to the following questions:

From your perspective, what are generative AI technologies?

Have you used generative AI in your teaching practices? If so, why and in what con- text? Provide examples.

From your perspective, what are the benefits of generative AI for adult EAL students? (p.6)

The analysis of responses to these questions demonstrated that teachers recognize the potential of generative AI (e.g., ChatGPT, Copilot) to enhance language learning and teaching. However, they also express concerns about when, where, and how it should be used, particularly regarding authenticity, integrity, and ethical implications.

The article discusses these findings through the lens of a framework of entangled pedagogy and teacher agency, Fawns (2022) in which technology, pedagogy, teacher agency and context mutually shape one another and that the relationship of pedagogy and technology is better understood by recognising the complexities of relational and contextual factors.

The authors argue that teacher agency is a “key concept in understanding how teachers and students interact with generative AI within learning and teaching contexts, particularly in how their individual capacities or capabilities (e.g. knowledge of generative AI), and relational and contextual factors (e.g. relationships teachers and students and/or access to generative AI) might influence their engagement with generative AI.” (p.5)

Although teachers acknowledged their agency to use AI responsibly, and their intentionality in using generative AI to create resources and to support learning, many chose to conceal their use of it due to relational and contextual pressures. The findings highlight the tension between AI’s promise to empower teachers and the challenges it poses to professional identity, pedagogical integrity, and trust in language education.

Retrievable from: https://www.tandfonline.com/doi/full/10.1080/09500782.2025.2485935#abstract

Environmental Scan: Application of AI in Settlement Services

Toronto East Quadrant Local Immigration Partnership (2024). Environmental Scan: Application of AI in Settlement Services. (Carolina Berinstein and John Saunders, CB Consulting).

This article reports the findings of an environmental scan on the application of AI (Artificial Intelligence) within the Toronto East Quadrant Local Immigration Partnership. Although the findings are not representative of the whole sector, this report offers a timely snapshot of current use of AI in this particular region as well as a discussion of the opportunities and challenges of AI, and a preliminary discussion of the development of guidelines and policies (with some examples) for the use of AI in settlement organizations.

Based on a literature review, a survey of members of the Toronto East Quadrant Local Immigration Partnership, and key informant interviews the report shows a growing interest in exploring and implementing AI in all aspects of settlement services (including language teaching and learning).

AI tools are being developed for services in the settlement sector including integration support, employment-related skill training and overcoming language barriers.  Respondents whose organization are using AI reported using generative AI tools such as ChatGPT, Gemini (from Google), Copilot (from Microsoft) and Google Translate, among others. One respondent reported the use of AI for class activities and lesson plans.

While interest in and use of AI is growing organizations in the settlement sector, as it is generally in Canada (According to the report of a survey of 2000 Canadians released in July 2025 by CIRA (Canadian Internet Registration Authority), AI use by Canadians has increased from 16%  in 2024 to 33% in 2025),  to date there are few examples of fully developed guidelines and policies for the use of AI in settlement organizations. 

This report includes a comprehensive discussion of the need for such guidelines and polices in the sector to support the ethical use of AI, and provides links to guidelines and policies developed by organization in the Canadian non-profit sector.

Retrievable from: https://accesemployment.ca/wp-content/uploads/2025/02/Environmental-Scan_-Application-of-AI-in-Settlement-Services.pdf

Integrating ChatGPT for vocabulary learning and retention: A classroom-based study of Saudi EFL learners

Abdelhalim, S. (2025). Integrating ChatGPT for vocabulary learning and retention: A classroom-based study of Saudi EFL learners. Language Learning & Technology, 29(1), 1-24.

This article provides the findings from a research project examining the use of generative AI (Chat GPT) to support vocabulary development in an EFL program in Saudi Arabia.  The research is grounded in sociocultural theory in which learning is seen as a socially mediated process through interaction with peers or tools within meaningful social contexts and in carrying out purposeful tasks. A key concept in sociocultural theory is the Zone of Proximal Development which proposes a range between what learners can do independently and what they can achieve with guidance. As the authors describe it,  

Effective scaffolding—structured, responsive support that gradually fades—helps learners move from assisted to autonomous performance (Schumm, 2006) p.12.

 They contend that AI tools including ChatGPT can be seen to function as dynamic mediators to support vocabulary development through real-time interaction, feedback and engagement with generative language tasks.

Although this classroom-based research study was conducted in an institutional setting and with a particular group of EFL students the quantitative and qualitative data indicate the potential of thoughtfully and carefully integrated generative AI to support vocabulary development across a range of language learning settings. In particular the qualitative data show that students found that ChatGPT was engaging, and reported that its use enhanced motivation, learner interaction and autonomous learning.

The researchers considered the following research questions:

  1. How does integrating ChatGPT into classroom instruction influence both receptive and productive vocabulary acquisition among EFL students compared to conventional teaching methods?
  2.  How sustainable are the vocabulary gains from ChatGPT-enhanced instruction compared to conventional teaching over time, as evidenced by delayed post-test results?
  3. What are students’ perceptions and evaluations of ChatGPT as a vocabulary learning tool in comparison to traditional classroom instruction? p.5

The researchers worked with two groups of students enrolled in one course with a single instructor. The first group used ChatGPT as an interactive tool for vocabulary practice and completed AI-generated exercises which were tailored to their progress. The second group engaged in traditional teacher-led learning. Quantitative results showed that the group using ChatGPT outperformed the students in the traditional classroom in productive vocabulary knowledge and in overall test scores. Researchers conducted pretests, and immediate and delayed post-tests as well as semi-structured interviews with students. Qualitative findings indicate that the use of generative AI promoted increased learning motivation and learner interaction. Students emphasized the ongoing need for instructor support and guidance, provision of structured prompt training and for feedback regarding AI outputs.

Overall, the findings indicate the potential of generative AI in supporting vocabulary development and point to the critical role of the language instructor in scaffolding learning, modelling how to create and use prompts, facilitating discussion around AI outputs and the critical interpretation of feedback. The authors stress the need for ongoing AI literacy, including digital and ethical training to enable students to critically assess resources, verify information, and to critically evaluate AI.

Over-reliance on AI, uncritical acceptance of generated content, and exposure to potentially inaccurate or culturally inappropriate examples underscores the need for AI literacy. p.13

Teachers also play a key role in validating AI-generated content and encouraging learners to question and refine it rather than accept it passively. p.13.

Retrievable from:

https://www.lltjournal.org/item/10125-73635/

Technostress and English language teaching in the age of generative AI

Kohnke, L., Zou, D., & Moorhouse, B. L. (2024). Technostress and English language teaching in the age of generative AI.
Educational Technology & Society, 27(2), 306-320.

This article offers a description, examination and discussion of “technostress” in the field of language instruction.  The authors describe technostress as “…a current-day affliction among  teachers that impacts their ability to adjust and respond healthily to the increased use of innovative technologies” (p.307).

The context of the study is EFL instruction in Hong Kong, in relation to the integration of Artificial Intelligence (AI) in the language learning field. However, the interview responses of the participant instructors, and the discussion of the factors that influence technostress and strategies to alleviate its impact have significant relevance for the ESL community, including instructors, programs and curriculum designers, who are all likely to encounter similar challenges in relation to the integration of a range of digital technologies, including AI tools such as ChatGPT in language instruction.

The authors outline five factors that influence technostress as described by Tarafdar (2019). Briefly, these are: Techno-complexity (the need to constantly learn how to use new technologies), Techno-overload (the proliferation of technology causes us to modify our work habits etc.), Techno-invasion (the need to use personal time to learn about new technologies etc.), Techno-insecurity (anxiety that new technologies will lead to job loss), Techno-uncertainty (uneasiness due to the frequent changes and updates to technology tools, etc.).

Based on the analysis of the qualitative date in this study and a review of the literature, the authors discuss strategies to alleviate technostress. These include the following:

Enabling instructors to develop or enhance Technological Pedagogical Content Knowledge (TPACK) to support effective integration of technology. 

Online Engagement, whereby instructors have opportunities to share knowledge about the integration and use of technology with their peers and colleagues in collaborative communities of practice.

Gradual and realistic expectations, whereby the integration of new technologies is incremental.  

Institutional support and clear policies and incentives so that instructors have a clearer understanding of expectations and access to professional learning opportunities.

Retrievable from:

https://www.researchgate.net/publication/378774496_Technostress_and_English_language_teaching_in_the_age_of_generative_AI

The future of language education: Teachers’ perceptions about the surge of AI writing tools

Zimotti, G., Frances, C., & Whitaker, L. (2024). The future of language education: Teachers’ perceptions about the surge of AI writing tools. Technology in Language Teaching & Learning, 6(2), 1–24.

This article reports on the results of a survey of language educators to explore their perceptions of the use and the potential effect of large language models such as ChatGPT in language teaching and learning. The researchers set out to address two questions:

RQ1: What current attitudes and perceptions do L2 instructors have regarding ChatGPT?

RQ2: What are innovative approaches for employing ChatGPT within language classrooms, their perceived ability to notice students’ use, and their current policies regarding this technology (p.3)?

In general, the survey revealed that language educators are open to the use of ChatGPT, recognise that their students will use it and that 36% of respondents have already used ChatGPT in their classrooms. Potential uses identified by respondents include creating lesson content, providing feedback, generating ideas and explaining complex grammar topics. Respondents also expressed concerns about the potential misuse of ChatGPT by students and how educators can identify and respond to misuse.

The findings align with the theory of normalization as described by Bax, in that educators who have used ChatGPT are enthusiastic about its potential, while those who have had less experience remain somewhat skeptical and apprehensive about its uses.

The authors argue that the inevitability of  the use of tools such as ChatGPT by language learners means that language educators need to critically examine such tools, focusing on the potential to enhance learning rather than on the technology itself, consider how these tools can be used for effective  student assessment and support students to think critically about these tools while enabling their responsible use to support the development of digital literacy skills. The article concludes with a quotation from a survey respondent, “We can’t change the fact that ChatGPT is going to change language teaching… all we can do is adjust our sails.” (p.20).

Retrievable from: https://www.castledown.com/journals/tltl/article/view/1136

AI and English language teaching: Affordances and challenges

Crompton, H., Edmett, A., Ichaporia, N., & Burke, D. (2024). AI and English language teaching: Affordances and challenges. British Journal of Educational Technology, 00, 1–27. https://doi.org/10.1111/bjet.13460

This article provides a systematic literature review focused on how Artificial Intelligence (AI) is currently being used in English Language teaching and learning. Based on this review the article also provides an analysis and discussion of affordances of AI, and identifies potential  limitations and challenges presented by AI in English language teaching and learning. The authors outline themes and issues requiring future research to support  the effective use of AI for English language teaching and learning at all learning levels.

The literature review demonstrated that AI is currently more likely to be used to support the learning in relation to speaking (primarily pronunciation) and writing  (primarily vocabulary and grammar) skills. The literature review indicated that AI is used less  in relation to reading,  and the authors posit that this may be due to the considerable  affordances of natural language processing which are  more pertinent in the areas of speaking and writing. The review also uncovered the potential of AI- powered tools including chatbots, on platforms such as Duolingo, Memrise and Mondly, to support learners in self-regulation, to become more autonomous learners, and to reduce anxiety, and increase confidence  in speaking English.

Based on their analysis of the literature the authors conclude that although there is solid evidence of promising affordances of AI, further research is required to better understand the implications of the use of AI in English language instruction. They contend that there is a  clear gap in the research in relation to challenges and limitations of AI, and identified the following areas for future research: technology breakdowns, concerns relating to personal cyber security, variable quality of AI-powered tools, and a potential over-standardisation of language.

Finally ,the authors point to the critical importance of enabling and supporting English language instructors to understand AI, how to evaluate AI-powered tools, and how to make effective use of these tools for language teaching and learning.

Retrievable from: https://bera-journals.onlinelibrary.wiley.com/doi/full/10.1111/bjet.13460

Agora World: A Glimpse into the Future of Technology in Language Teaching and Learning

Cowin, J. (2023) Agora World: A Glimpse into the Future of Technology in Language Teaching and Learning. NYS TESOL Journal 10 (1), 51-58.

This article provides an intriguing “glimpse” into the future of language learning and the potential of virtual reality (VR) and multimedia technologies to transform language education. Through a materials review of Agora World  a code-free, drag and drop platform which allows users to fairly easily build immersive experiences, the author explores its potential to enable language instructors to create virtual worlds and immersive experiences in which learners can engage in meaningful tasks and activities to support and enhance their language learning. “Agora World holds the promise of content-based, project-based, or task-based language learning, making it possible for language teachers to construct original virtual worlds, gallery walks, and learning pathways for L2 learners to master language production.”.

These tasks and activities could include gallery walks in which learners can practice conversations, quests in which learners can carry out assigned tasks , including solving a problem together, and experiment with new vocabulary.  In each case instructors can  review and monitor progress and participation. Learners can also create stories and media boards to show language concepts learned, and these can be included in portfolios for assessment.

The author notes that while VR and immersive technologies hold much promise for language learning, the role of the instructor will be crucial in  designing and creating and curating appropriate and meaningful immersive content to support language learners.

Agora World provides a free version with most of the functionality of a subscription version which enables users to explore and evaluate the potential for teaching and learning.

Retrievable from: https://journal.nystesol.org/Vol10no1/Cowin_V10_1.pdf

Potential Affordances of Generative AI in Language Education: Demonstrations and an Evaluative Framework

Pack, A. & Maloney, J. (2023). Potential Affordance of Generative AI in Language Education: Demonstrations and an Evaluative Framework. Teaching English with Technology 23(2),4-24.

This timely article provides a comprehensive overview of the potential affordances of Artificial Intelligence (AI) in language teaching, particularly the AI chatbot ChatGPT. As the authors note, much of the current discourse on AI in the education sector is concerned with how AI can or could be misused by students. Although serious issues such as plagiarism need attention, the article focuses on the potential of tools such as ChatGPT to support teaching and learning and to save valuable time in developing learning materials, assessment tasks and rubrics for evaluating student writing.

AI chatbots such as ChatGPT work through prompts provided by the users, and the authors offer examples of prompts designed to elicit quality, appropriate responses which can aid instructors in developing materials for instruction and learning and for assessment purposes. The examples provided are based on the Common European Framework of Reference for Languages (CEFR) but could be drawn from other frameworks such as the CLB framework and/ or the LINC curriculum guidelines. The article also includes a discussion of how two existing frameworks, Hubbard‘s (1988, 2021) framework for evaluation and Bronfenbrenner‘s(1979), can be used in combination to assist language instructors in making informed decisions about when, how and why to use generative AI to support materials development and the creation of assessment tasks to be used in their own context.

The authors suggest that by combining elements of each framework, instructors can evaluate the use of AI tools for themselves and their students and consider how their use will align with the values and policies of the institutions and programs in which they work.

Retrievable from: https://files.eric.ed.gov/fulltext/EJ1397173.pdf

ChatGPT for Language Teaching and Learning

 Kohnke, L ., Benjamin  Moorhouse, L. & Di Zou, D. (2023) ChatGPT for Language Teaching and Learning. RELC Journal OnlineFirst, April 3, 2023.

 This short article provides a useful introductory technology review and discussion of the possible affordances and the challenges of the AI chatbot ChatGPT for language teaching and learning. The article outlines some of the possible affordances of Chat GPT, for instructors and students. For example, ChatGPT can be used to identify word meanings in context, explain language mistakes, generate texts, e.g., emails and stories etc., develop quizzes and provide dictionary definitions.  

There are several ongoing debates about the use of ChatGPT in education generally.  Educators have questioned the ethical use of ChatGPT including the possibility of cheating, and how it may affect assessment. Since it does not supply sources or citations, there are concerns about the accuracy of ChatGPT responses which may mislead students. There are also questions about cultural bias, since most of the texts in the database are English; this may be of particular concern in language education as students from diverse cultural backgrounds may not be aware that this tool is not culturally neutral.

The authors argue that the effective use of ChatGPT will require the development of specific digital competencies in three broad categories. Instructors will need Technological Proficiency, including understanding the features of ChatGPT and how it works; Pedagogical Compatibility, that is consideration of how this tool could be used to enhance language teaching and learning, planning for implementation, and providing guidance for students to use the tool for independent learning. Social Awareness includes critical awareness of the challenges of ChatGPT and supporting students to understand the challenges and the risks and ethical issues inherent in the use of the tool.

The article concludes with a list of suggestions of how students can use ChatGPT to improve their English language learning.

Retrievable from:

https://journals.sagepub.com/doi/epub/10.1177/00336882231162868

The Affordances and Contradictions of AI-Generated Text for Second Language Writers

Warschauer, M., et al. (2023). The Affordances and Contradictions of AI-Generated Text for Second Language Writers SSRN.

This timely report looks at how to approach the affordances of AI-writing tools like ChatGPT in second language classrooms, while maintaining the importance of learners first developing writing skills without these tools. The authors encourage ongoing critical discussion about AI-generated text to ensure that educators learn how to manage it in their educational context, rather than simply banning it. They make the point that learners will most likely be expected to be able to use AI-generated text in employment situations and will be at a disadvantage if they have no experience with the tools or think of them as “cheating tools” or shortcuts.

The authors propose a five-part framework to help learners find their way with ChatGPT. It includes these components: understand, access, prompt, corroborate and incorporate. They provide a list of prompt types and examples of each one. They describe and explain the reasons for each of the other components of the framework. The report also presents a sample list of the functions of ChatGPT to explain some of its uses.

Retrievable from:

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4404380

Intelligent assistants in language learning: an analysis of features and limitations

Kukulska-Hulme, A. & Lee, H. “Intelligent assistants in language learning: an analysis of features and limitations.” In CALL for widening participation: short papers from EUROCALL 2020, edited by Karen-Margrete Frederiksen, Sanne Larsen, Linda Bradley and Sylvie Thouësny, 172-176 .2020

This short article presents the findings of a multidisciplinary literature review of research studies conducted over the past 10 years examining the development and potential applications of Intelligent Assistants, including chatbots, avatars and conversational agents such as Siri and Alexa, in the field of education.

The article focusses on the features and limitations of Intelligent Assistants to support language learning. Based on their analysis of over 170 studies the authors find that Intelligent Assistants have the potential to support language learners, at all learning levels. Using an Intelligent Assistant provides learners with opportunities for extensive simple conversation and pronunciation practice, anytime and place, and, since the conversation is taking place privately, can help reduce the anxiety that many learners experience when attempting to speak in a new language.

The authors note that research is needed to learn more about how Intelligent Assistants can be effectively deployed in the language classroom and what the role of the instructor could be in supporting learners to make best use of these tools.

To see a chatbot in action visit: http://www.costi.org/  and click on the icon at the bottom right.  The Orientation to Ontario chatbot provides basic information about living and working in Ontario. The chatbot allows users to chat in text, and to input questions verbally.

Retrievable from: https://files.eric.ed.gov/fulltext/ED611082.pdf

Goodbye to classroom teaching? ARTIFICIAL INTELLIGENCE IN LANGUAGE LEARNING

Lotze, N. Goodbye to classroom teaching? Artificial Intelligence in Language Learning. (2018). Magazin Sprache. May.

This article offers a critical response to claims that are made about how Artificial Intelligence (AI) can profoundly alter the ways in which languages can be taught and learned, i.e.,  without the necessity for classroom teaching and learning.

The author makes a succinct and quite compelling argument that such technological innovation actually provides quite a limited range of potential for language learning. For example digital exercises, although they may appear innovative, are actually replications of what can be found in traditional exercise books, the usefulness of learning language apps with a chatbot, (Duolingo for example) is limited, in that such systems depend on simple, defined interactions and predictable dialogues.

The author concludes that learning language systems based on AI are useful when it may be difficult or impossible for students to attend language classes, and are best suited for beginning language learners to supplement classroom learning with a human teacher.

Retrievable from: https://www.goethe.de/en/spr/eng/gls/21290629.html