Generative AI tools and empowerment in L2 academic writing

Generative Artificial Intelligence (GenAI) developments have ignited interest in how they will impact L2 writers' agency over their academic writing processes. At the same time, higher education (HE) institutions have responded to GenAI by devising policies on their use. Recognizing that L2 writers' processes can be shaped by GenAI, as well as their institutions and instructors' policies, there is a need to understand how L2 writers engage with GenAI tools and negotiate agency as they transition into their studies and find themselves within the boundaries of their HE institution's policies. Through the lens of critical digital literacies (CDL), agency and empowerment, this qualitative study adopted in-depth focus group interviews with a preference selection task to explore how twenty-one post-graduate L2 writers position themselves in relation to GenAI tools and how they and GenAI have been positioned by their instructors. Data were collected within the first month of their post-graduate studies. The findings show that the L2 writers have integrated GenAI into various aspects of their academic writing processes. Their responses suggest that the L2 writers believe the tools empower them and augment their writing abilities. Yet, from a CDL perspective, this may be false empowerment, as the writers showed little critical awareness of how the tools work. At the same time, tensions exist between their perceptions of agency and some of their instructors' GenAI policies. The L2 writers prefer GenAI policies that give them the autonomy to use GenAI flexibly. However, they recognize the need to be accountable for their academic writing and transparent in their GenAI use. Bans on GenAI use were perceived as unfair, restricted agency, and not reflective of GenAI affordances. The findings can inform academic writing training and instructors' policies.