This study explores how international students from the Western Balkans use artificial intelligence (AI) tools to support academic adaptation, writing development and self-directed learning in Anglophone higher education institutions. By examining their lived experiences through qualitative interviews, this research aims to understand AI’s role as both a cognitive scaffold and a transformational partner in learning. This study also addresses equity, ethics and overreliance, contributing theoretical insights and practical implications for inclusive, AI-enhanced education.
The study used a qualitative research design within a constructivist–interpretivist paradigm to explore international students’ experiences with AI. Nine undergraduate students from Kosovo, Albania and North Macedonia studying at UK and US institutions were recruited through purposive sampling. Data were collected in three phases: orientation, semi-structured interviews in Albanian and member checking. Interviews were transcribed, translated and analyzed using Braun and Clarke’s (2006) six-phase thematic analysis. Intercoder agreement, reflexive journaling and an audit trail enhanced rigor. Ethical standards were followed in line with the Declaration of Helsinki and APA guidelines.
AI tools supported international students’ adaptation by simplifying complex texts, clarifying assignments and reducing language-related anxiety. Participants described AI as a “silent tutor,” enhancing comprehension, academic writing and confidence. Beyond correction, AI fostered self-directed learning by enabling students to plan, summarize and test understanding independently. However, concerns emerged around unequal access to premium tools, ethical boundaries and potential overreliance, which risk undermining critical thinking and autonomy. Overall, AI was experienced as both a scaffold for inclusion and a challenge requiring institutional guidance to ensure equitable and responsible use in higher education.
The study’s small, purposive sample of nine students from the Western Balkans limits generalizability and may not capture the full diversity of international learners. Reliance on self-reported experiences also introduces subjectivity. Despite these constraints, the findings provide valuable insights into an underexplored population in AI-in-education research. Future studies should use larger, comparative and longitudinal designs to examine evolving patterns of AI use across contexts. Expanding beyond Western Balkan students will further validate and extend these findings, helping to shape inclusive, ethical and evidence-based policies for AI integration in higher education.
The findings highlight the need for universities to integrate AI literacy into academic skills programs, enabling students to use tools ethically and effectively. Faculty can leverage AI to scaffold writing, comprehension and independent learning while modeling responsible practices. Institutions should provide equitable access to premium AI platforms to reduce disparities among students from different economic backgrounds. Clear policies and support structures are essential to balance innovation with integrity, ensuring AI use enhances, rather than undermines, autonomy and critical thinking. Such measures can foster inclusive, AI-enhanced learning environments that better support international students.
This study underscores how AI can reduce barriers for culturally and linguistically diverse learners, supporting inclusion and academic success in global higher education. At the same time, unequal access to premium AI tools risks reinforcing social and digital divides, particularly for students from less-resourced regions. Establishing equitable access and ethical frameworks is therefore critical to ensure AI empowers rather than marginalizes learners. By addressing these issues, universities and policymakers can promote fairness, strengthen international student integration and foster a more inclusive, technologically supported society where AI contributes to educational opportunity and social mobility.
To the best of the authors’ knowledge, this study is one of the first to explore how international students from the Western Balkans engage with AI to support academic adaptation, writing and self-directed learning in Anglophone higher education. By focusing on an underrepresented student population, it offers fresh insights into the cultural and linguistic dimensions of AI use, beyond commonly studied Global North and Asian contexts. The integration of transformational learning (Mezirow, 1997) and self-directed learning (Knowles, 1975) theories provides a novel interpretive lens, extending their application into digital, AI-mediated environments and contributing both theoretical depth and practical guidance for inclusive AI adoption.
