Abstract
Generative AI is rapidly transforming higher education by reshaping cognitive processes, learning behaviors, assessment practices, and instructional approaches. This study examines the impact of AI on student learning through a combination of multi-institutional evidence and a quasi-experimental assessment in an undergraduate writing course. Three central dimensions are analyzed: cognitive offloading, critical versus naïve adoption of AI, and emerging learning patterns including normalization, confirmation bias, and the erosion of scaffolding. Findings reveal that AI tools can enhance grammar accuracy, research efficiency, and factual recall, while also posing risks to creativity, critical thinking, independent revision, and metacognitive engagement. The study highlights the importance of structured, critically mediated integration of AI into curricula to maximize learning benefits, uphold academic integrity, and support long-term skill development.
Keywords