2026 ASEE Annual Conference & Exposition

This complete research paper reports the extent of prior use (in high school), planned use in first-year (FY) engineering courses, and disclosed use of Generative AI (GenAI) for various tasks in two first-year engineering courses at a single large public institution. Given the rapid evolution of GenAI it is important for instructors of FY engineering courses to understand the previous use history and plans of their students to facilitate appropriate use and discourage uses of GenAI that constitute cheating. It is also important to avoid exacerbating or creating achievement and learning gaps among students from different demographic groups. The research was conducted in the fall 2025 with over 500 students enrolled in a first-year engineering projects (FYEP) course with open-ended team design projects or an engineering tools and analysis course that had learning objectives related to math skills and Matlab. The majority of the students had used GenAI in schoolwork prior to college, with higher use among engineering students compared to students exploring engineering as a major. Disclosed previous GenAI use was higher among male compared to female students. Planned GenAI use varied among types of course activities, with the highest frequency for exploring unfamiliar topics and lowest for help with writing. Planned use of GenAI for some activities was higher among male compared to female students and higher among students who first language was not English. Some differences in planned GenAI use were also found among White, Hispanic, and Asian students. Planned use was weakly correlated with students’ reported use at the end of the semester in the FYEP course. On average, reported student use of GenAI during the semester was higher in the engineering problem solving course compared to the FYEP course. Substantial section level variation in GenAI use and policy perceptions was observed within the first year engineering projects course, with uneven student understanding of policies even within the same section, highlighting challenges of policy clarity and consistency in multi section courses. Taken together, these findings show that first year students enter engineering with prior experience using GenAI, develop intentions for use early in the semester, and adapt their practices in response to course context and instructor policies. As GenAI tools become increasingly embedded in academic and professional contexts, early and explicit guidance in first year courses will be essential for fostering responsible use while mitigating potential negative impacts on learning and equity.