Impact of Generative AI on Student Learning Behavior: Evidence, Patterns, and Educational Implications

Abstract

Generative Artificial Intelligence (GenAI) has already become a fast solution in the student workflow and changed the way learners search, draft, revise, and practice academic content. Among the short-term issues of academic misconduct, one of the key educational questions is how GenAI is changing the learning behavior, specifically, effort distribution, self-regulation, verification practices, and conceptual knowledge. This paper analyzes the impact of GenAI on learning behavior of students through a structured questionnaire with short guided reflections. The article includes the trends of GenAI application on typical academic activities (concept explanation, summarization, writing support, and coding assistance), the changes in students’ behavioral patterns in studying, and the differences between learning-supportive and learning-substitutive application. Results show that GenAI is capable of facilitating the learning process when it is applied as a scaffold of explanation, feedback, and self-testing by students, and when it is regularly checked against reliable sources. Nevertheless, the quality of learning can decrease when the students use GenAI as the main answer-generator, revise less, and verify less. The findings indicate that task design, assessment model, and AI literacy do influence responsible use. At the end of the paper, recommendations are provided to educate and institutions that intend to make the most out of GenAI and minimize its risks to learning and academic quality.