Engineering Problems have long been used in the classroom for educational purposes. The process ofgenerating said problems has not changed significantly in decades. However, the advent of large language models presents an opportunity to explore how such AI tools may be used to support problem generation. This thesis proposes a novel approach to redefine the generation of engineering problems using Generative Artificial Intelligence (Gen-AI) and acknowledging its unique features. Students’ preferences, performance, mental workload, and emotions are evaluated using a mixed-method research approach with different generation sources. Ultimately, the study identifies substantial impacts of the generation of problems on students. The insights of this research contribute to developing the landscape of engineering pedagogy through the potential of Gen-AI to redefine traditional engineering problems and improve students’ overall learning experiences.