Empowering International PhD Students: Generative AI, Ubuntu, and the Decolonisation of Academic Communication

Generative AI (GenAI) has several affordances that makes it valuable for learning. However, there are also several challenges, most notably ethical considerations. While much of the current research on GenAI in academia focuses on technical capabilities or ethical concerns, few studies have examined how GenAI can be leveraged to promote equity and inclusivity, particularly in academic communication. In this study, we adopt the theoretical lens of Ubuntu, a Southern African philosophy that emphasises interconnectedness and community, to reconceptualise GenAI as a democratising force. Five international PhD students from Education and their PhD supervisor (all multilingual users of English) explored their experiences of using GenAI. We used a collaborative autoethnographic approach, incorporating reflective prompting, individual writing, group sharing and refining, and group-based writing. We demonstrate that GenAI reshaped power dynamics and challenged academic hierarchies by providing real-time language support and improving academic writing clarity. This empowered us to participate in academic discourse, disrupting traditional gatekeeping mechanisms. By automating routine tasks, GenAI also shifted our academic focus from technical skills to intellectual contributions, fostering inclusivity and equity. This study highlights the potential of GenAI in terms of knowledge creation and academic discourse, particularly for doctoral scholars navigating linguistic and cultural barriers. Academia is predominantly Anglophone-centric, marginalising non-English-speaking scholars by imposing additional barriers to participation and recognition, perpetuating inequities in knowledge creation and dissemination. We argue that embracing GenAI can help to decolonise these academic practices, thereby promoting epistemic justice and creating a more equitable scholarly environment for all.

This record includes the full paper as well as two datasets which are all made available for use by other researchers under the CC-BY 4.0 licence. This paper and the associated data for the project have been made available via pre-print for two key reasons. Firstly, given the rapidly evolving landscape of generative AI research, we aimed to share our findings promptly, allowing other scholars to build on our work without delay. Secondly, in alignment with our commitment to democratising knowledge creation and dissemination in academia, we sought to make our research as widely accessible as possible to foster inclusivity and collaboration across the scholarly community.