In the context of internationalized higher education, students increasingly turn to generative AI tools like ChatGPT to support academic writing and translation. This presentation investigates how students independently engage with ChatGPT outside formal classroom settings, focusing on the everyday practices, motivations, and ethical ambiguities that shape their use. Drawing from quantitative data this study explores how students use ChatGPT to draft essays, translate academic texts, and refine language for scholarly communication. While many view the tool as an accessible linguistic ally in navigating the demands of academic English or other target languages, their practices also raise critical questions about authorship, academic integrity, and linguistic agency. The talk critically examines how these emergent AI-mediated literacies intersect with broader themes of intercultural communication, educational inequality, and institutional responsibility in a globalized academic landscape. It invites reflection on how universities might rethink support structures, language policies, and ethical frameworks in response to this evolving terrain.

--