Exploring the influences of socio-economic background, confidence, and ethical awareness on students' use of generative artificial intelligence (GenAI) in mathematics education

This study explored the influence of socio-economic background, mathematical confidence, and ethical awareness on students' use of generative artificial intelligence (GenAI) in mathematics learning at a Canadian college serving international and working students. Guided by a phenomenological orientation and the Presage-Process-Product framework, nineteen students were interviewed and the data analyzed thematically using NVivo 14. Automation - using GenAI primarily for task completion - was more common than augmentation, in which AI supported understanding; self-regulated and ethically regulated uses were also observed. Socio-economic pressures shaped when and how students turned to AI, with many shifting between augmentation and automation under work and deadline pressure. Confidence was related to how, rather than whether, students used GenAI, and was reshaped by use in turn. Students expressed concerns about inaccuracies and overreliance, but awareness of socio-ecological costs remained uneven. The study distinguishes strategic reliance from dependence and calls for responsible, ethically informed GenAI use.