Context

The research is prompted by a classroom observation during a seminar in which several students were wearing AI-enabled smart glasses to translate spoken English into written text in their language real time. These devices record lectures, translate them and upload data to cloud-based systems, over which lecturers have no control, raising concerns about data ownership, consent, and academic governance. This moment prompted critical reflection on how emerging translation technologies are transforming international postgraduate learning environments.

The study is situated within a postgraduate programme at a Scottish university in which over 98% of students are international. In this highly multilingual context, the increasing availability of AI has led to growing student reliance on translation tools to support lecture comprehension, seminar participation, and academic writing. Classroom observations indicate a significant shift in how linguistic accessibility is mediated, with AI technologies increasingly functioning as intermediaries between students and lecturers.

Problem

Although AI translation technologies may enhance access and inclusivity by lowering linguistic barriers (Fitas, 2025), their educational implications remain underexamined. Existing research suggests that reliance on translation tools may limit opportunities for language development and authentic interaction (Zhang, Li & Wu, 2025; Rapa et al., 2024), and inaccuracies in translation can result in partial or distorted comprehension (Altakhaineh et al., 2025).

Moreover, the use of AI translation tools raises ethical concerns related to data privacy, algorithmic bias, and institutional governance (Regan & Jesse, 2019). In particular, the recording and cloud storage of lecturers’ speech may expose sensitive academic content and personal data to potential misuse. These issues are compounded by the fact that institutional AI policies tend to prioritise generative AI, often overlooking translation technologies, leaving gaps in guidance for both staff and students.

Aims and research questions

This project investigates the use of AI translation tools by international students in live EMI teaching sessions in a postgraduate programme at a Scottish university. The study aims to examine how these tools are used in practice and analyse their impact on students’ comprehension, engagement, and academic confidence.

Specifically, the study seeks to explore students’ motivations for using these tools during lectures and seminars, the ways in which they are integrated into live learning activities, and students’ perceptions of their benefits and limitations.

By systematically analysing cloud storage practices underpinning commonly used AI tools and their pedagogical effects, the project aims to deliver evidence-based guidance for institutions seeking to implement responsible and inclusive AI in multilingual higher education contexts.

The study addresses the following research questions:

1. What are the primary motivations behind international students’ use of real-time AI translation during live lectures and seminar/tutorial discussions?

2. How do they use these tools and what specific tools do they use?

3. What benefits and challenges do students experience when using AI translation tools in EMI contexts?

Preliminary analysis of existing literature indicates that, although a growing body of research has examined the role of machine translation in educational settings, limited attention has been paid to the use of AI-based translation tools within English-medium instruction (EMI) contexts. There is a lack of focused empirical and conceptual work examining how AI translation tools affect academic integrity and mediate the linguistic demands experienced by both students and instructors in EMI environments. This gap underscores the need for targeted research to clarify the pedagogical, ethical, and linguistic implications of AI translation in these settings.

This study employs a mixed-methods research design informed by a pragmatic paradigm. The study is exploratory prioritising enhancement of pedagogical practice while contributing to broader educational debates (Fanghanel et al., 2016). Technology Acceptance Model (TAM) (Davis, 1989) provides the theoretical framework, offering a well-established lens through which to analyse students’ adoption of AI translation tools, particularly in relation to perceived usefulness and ease of use. Data collection included an online survey distributed to 200 international postgraduate students, who speak English as a foreign language, enrolled in an education programme;168 students completed the survey. In-depth semi-structured interviews were conducted with 12 participants selected from survey respondents. Recruitment took place via email and course announcements, with participation voluntary and withdrawal permitted at any stage prior to data analysis. To ensure diversity, participants from a range of linguistic and cultural backgrounds who speak English as a foreign language were contacted. Interview participants were selected through purposive sampling, targeting students who reported active use of AI translation tools in academic contexts. Semi-structured interviews, approximately 30-45 minutes in duration, were conducted either online or in person according to participant preference. Data collection began with an anonymous online survey designed to capture types of AI translation tools used, frequency and contexts of use, and self-reported benefits and challenges. Survey invited interested participants to contact the researchers if they wished to be interviewed. Semi-structured interviews with 12 participants were subsequently used to explore motivations for tool use, impacts on classroom participation, and ethical considerations in greater depth. Quantitative survey data were analysed descriptively using Qualtrics to identify trends and patterns. Qualitative interview data are being analysed thematically following Braun and Clarke’s (2006) six-phase framework. Emergent themes will be compared with existing literature to situate findings within broader pedagogical and technological debates. Ethical approval for the project was granted by the School's ethics committee. The study adheres to established ethical principles, including informed consent, confidentiality, voluntary participation, and secure data management in compliance with GDPR and institutional policy. Participants received comprehensive information sheets and were provided informed consent prior to participation.

The findings indicate that students’ use of AI-enabled translation tools is shaped not only by practical learning needs but also by affective and social considerations. Many participants reported downloading AI tools prior to moving to the UK after encountering discussions on local social media that characterised the Scottish accent as particularly challenging to understand. Many shared feeling self-conscious when using translation tools in class, expressing concern that their peers or lecturers might judge them as less capable. This sense of stigma was particularly pronounced in interactive classroom settings, where students felt that visible reliance on translation tools could undermine perceptions of their academic competence or language proficiency. As a result, some students described attempting to conceal their use of such tools despite recognising the potential benefits of real-time translation. A recurring theme in the data was uncertainty surrounding institutional policy. Students frequently expressed confusion about whether the use of translation tools was formally permitted during teaching sessions or in the completion of assessed work. In the absence of clear guidance, many adopted pragmatic strategies, relying on a combination of AI-powered translation, grammar-checking, and paraphrasing tools to support comprehension and note-taking. Several students also reflected on the longer-term effects of sustained translation tool use on their writing development. Some observed that their academic writing had begun to mirror the structure, tone, or phrasing commonly associated with AI-generated text, raising concerns about voice and originality. These reflections suggest that translation technologies may be influencing not only linguistic accuracy but also stylistic norms. Finally, the findings underscore significant equity and ethical issues linked to tool accessibility. Most students relied on free AI translation services, which were perceived as convenient but problematic. Participants reported concerns about data privacy, inconsistent recognition of diverse accents, and frequent translation inaccuracies, all of which affected comprehension and confidence.

Altakhaineh, A. R. M., Alghathian, G. A., and Jarrah, M. M. (2025). A comparative study of accuracy in human vs. AI translation of legal documents into Arabic. International Journal of Language & Law, 14, 63-80. https://doi.org/10.14762/jll.2025.063 Braun, V., and Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77–101. https://doi.org/10.1191/1478088706qp063oa Davis, F. D. (1989), Perceived usefulness, perceived ease of use, and user acceptance of information technology, MIS Quarterly, 13 (3): 319–340, https://doi.org/10.2307/249008 Fanghanel, J., Pritchard, J., Potter, J. and Wisker, G. (2016). Defining and Supporting the Scholarship of Teaching and Learning (SoTL). Higher Education Academy. Fitas, R. (2025). Inclusive education with AI: supporting special needs and tackling language barriers. AI Ethics. https://doi.org/10.1007/s43681-025-00824-3 Rapa, A.A., Saja, I., Azmi, A. (2024). The Use of Artificial Intelligence (AI) Translation Tools: Implications for Third Language Proficiency. International Journal of Research and Innovation in Social Science (IJRISS), 8(09), 1952-1960. https://dx.doi.org/10.47772/IJRISS.2024.8090161 Regan, P. M., and Jesse, J. (2019). Ethical Challenges of Edtech, Big Data and Personalized Learning: Twenty-First Century Student Sorting and Tracking. Ethics and Information Technology, 21, 167-179. https://doi.org/10.1007/s10676-018-9492-2 Zhang, W., Li, A.W. and Wu, C. (2025). University students’ perceptions of using generative AI in translation practices. Instr Sci 53, 633–655 https://doi.org/10.1007/s11251-025-09705-y