Generative Artificial Intelligence (GenAI) is increasingly visible inteacher education, with pre-service teachers (PSTs) using toolssuch as ChatGPT for lesson planning and workload management.Yet prior studies treat PSTs as a homogeneous group, overlookinghow year-group context and placement stress relate to perceptionsand uses of GenAI during placement. This study addresses this gapusing an explanatory sequential mixed-methods design, combiningsurvey and follow-up interviews. Quantitative data were collectedfrom 190 PSTs across four years of a Bachelor of Education pro-gramme in Ireland, and qualitative insights were drawn from eightstudents, two from each year. Findings reveal cohort-level profiles,where placement stress is associated with forms of GenAI engage-ment, linked to caution in Year 1 and Year 2 accounts and prag-matic GenAI use in Year 3 and Year 4 accounts. The study offersa cohort-based Stress – Adoption Reconfiguration Pattern and high-lights scaffolded, placement-sensitive GenAI preparation acrossinitial teacher education.