To succeed in higher academic education, students are expected to master the academic language, which can be particularly challenging for students whose first language is not Swedish (L2). AI technology offers customized opportunities to understand and produce academic texts, which can reduce cognitive load for multilingual students by allowing them to use multiple languages. By utilising the interactive features offered, students can receive support in understanding academic content and developing their writing skills. However, applying AI technology also comes with risks such as AI plagiarism. The expected contributions of this study are to increase understanding of how L2 students perceive and use AI technology resources in academic reading and writing within Swedish higher education. Such insights can show how AI technology can reduce language barriers and cognitive load, providing these students with better opportunities to focus on academic content. The study aims to analyse how the application of AI in academic literacy practices may reshape the conditions for students with Swedish as a second language in higher education. To reach this aim qualitative interviews were conducted with eight preschool teacher students with Swedish as their second language.

Each student participated in two different interviews:
(1) a semi-structured interview and
(2) a follow-up interview of a more narrative/reflective nature.

The interviews were analyzed thematically using both inductive and deductive approaches. Initially, an inductive analysis was conducted in five steps based on the study's research questions. Subsequently, a deductive analysis was performed, categorizing the interview material according to the concepts of affordances and student agency. Results indicate that L2 students perceive the use of AI in academic reading and writing comes with affordances such as; Language Support Opportunities, Content Understanding Opportunities and Equity Opportunities. The risks students perceived were linked to the following themes: cheating and false information, and lack of personal ownership of knowledge. Regarding student agency, results indicate that students' perceptions were linked to the following themes: opportunities for increased understanding and confidence, opportunities for flexibility in studies, and opportunities for increased independence. The conclusion is that AI technology has the potential to transform the conditions for various student groups, particularly for L2 students, which can lead to more equitable conditions compared to students who have Swedish as their first language. However, it is crucial to include education on the opportunities and risks that AI usage in higher education can entail to prevent dishonest use of AI that could lead to unintentional plagiarism.

Keywords: Academic Literacy, AI technology, Higher education, Second Language Students.