Academic Support or Misconduct? Chinese International Students' Experiences with Generative AI in UK Higher Education

Abstract

As Generative AI tools such as ChatGPT become more common in higher education, they are changing how students approach academic work and expectations. This shift raises questions about how agency, authorship and critical judgement are experienced in learning environments where Generative AI plays a role. Drawing on the framework of networked learning, which sees learning as resulting from interactions between people, resources and tools, this study explores how such interactions may be altered when students rely on Generative AI systems that operate outside of collaborative learning networks. Focusing on Chinese international students in the UK, who often face challenges related to language, academic norms and institutional trust, the research investigates how they use, interpret and respond to Generative AI tools in academic writing. Drawing on constructivism learning theory and posthumanism perspectives, it conceptualises student Generative AI relations not as a simple or straightforward process but as an evolving process influenced by tools, cultural expectations, and university policies. These theoretical lenses allow for a more critical and situated understanding of how Generative AI influence students' agency, perceptions of authorship and sense of ethical responsibility in academic work. The study adopts a three-stage design. First, an online survey explores the scope and purpose of students' use of Generative AI across various academic tasks. Second, a screen-recorded writing task observes how students interact with Generative AI in real-time during an authentic academic activity. Third, semi-structured interviews capture students' views on university rules, ethics and how they make choices. This design helps reveal both students’ visible writing choices and the less obvious thinking behind them. At the time of submission, data collection is still underway and will be completed by early 2026. Preliminary findings will be shared at the Networked Learning Conference to support broader discussions around student agency, academic integrity and ethical engagement with Generative AI in university settings shaped by digital tools. By focusing on the lived experiences of a specific student population, this research aims to contribute to ongoing theoretical and practical discussions on how learning, identity and authorship are changing.