Understanding the interaction between generative AI and factual knowledge in supporting students’ problem-solving

Generative Artificial Intelligence (GenAI) has the potential to transform how students approach cognitively demanding tasks by reducing cognitive load and supporting problem-solving. While this transformation may reshape the role of factual knowledge in curricula, little is known about how the benefits of GenAI depend on learners’ factual knowledge in cognitively demanding tasks. To address this gap, this study employed a mixed-methods design with 71 participants in a university-level web design course. Results showed that GenAI reduced cognitive load by providing scaffolding, lowering extraneous load, and optimizing intrinsic load, thereby making tasks more manageable. GenAI particularly improved task performance among students with lower factual knowledge, yet factual knowledge remained essential for critically evaluating, refining, and applying GenAI-generated outputs. This study demonstrates how GenAI interacts with learners’ factual knowledge to shape problem-solving, offering insights for advancing human–AI collaboration in education.