Generative AI (GenAI) is increasingly used in academic writing (Eleftheriou et al., 2025). Recent studies have examined how GenAI can be incorporated to support students at multiple stages of the writing process and to provide feedback (e.g. Allen & Mizumoto). While these tools offer promising opportunities, they also raise concerns and challenges, particularly regarding academic integrity. Despite growing research in this area, the role of human instructors in academic writing education in the era of GenAI remains underexplored. This study examines the uniqueness of human instruction by investigating the challenges that have emerged with the introduction of GenAI in academic writing and the strategies tutors use to address them during one-on-one sessions. Using a qualitative design, the study employs semi-structured interviews with eight tutors at an academic writing center at a Japanese university. Preliminary findings indicate that ethical concerns and a lack of clear guidance are the primary challenges tutors encounter when students use GenAI in their writing. Follow-up interviews will be conducted to identify emerging challenges and changes in instructional strategies over time. By foregrounding instructors’ experiences, this study aims to provide insights into future directions for human instruction in academic writing and more effective support for students.