"Heartbreak and Frustration'' experiences of healthcare providers caring for pregnant women with sickle cell disease in Uganda.

GenAI is rapidly reshaping teaching and learning in higher education. Among its emerging applications, AI writing tools such as ChatGPT have attracted particular attention because they can assist users with idea generation [], language correction [], text organization [], and revision across multiple stages of the writing process []. In L2 writing contexts, these tools are often seen as especially promising, as they may help learners reduce linguistic barriers, improve writing efficiency, and access immediate support during drafting and revising. Recent studies have shown that AI writing tools can offer important pedagogical affordances in L2 writing, including feedback support [], language scaffolding [], and enhanced learner engagement []. At the same time, however, their use has raised substantial concerns related to overreliance [], academic integrity [,], authorship [], and data privacy []. These concerns are particularly salient in educational settings, where students may simultaneously perceive AI tools as highly useful and potentially problematic. As a result, students’ adoption of AI writing tools cannot be understood solely in terms of technological convenience or instructional benefit. Although prior research has increasingly examined the educational use of GenAI, much of the existing literature has focused either on the pedagogical potential of AI writing tools [,,] or on ethical concerns surrounding their use [,,]. Comparatively less attention has been paid to how students negotiate these two dimensions at the same time when deciding whether and how to engage with AI writing tools in L2 writing-related tasks []. This tension points to a central but insufficiently examined issue: students may recognize the usefulness of AI writing tools while simultaneously becoming more aware of their risks.