AI and Study Habits among Generation Z: A Mixed-Methods Case Study of Undergraduate Learners at the University of Gujrat

The University of Gujrat (UoG) represents a large and diverse academic community composed mainly of Generation Z learners. Despite this, little research has explored how artificial-intelligence (AI) tools such as ChatGPT and Khanmigo influence students’ study habits. This study examined patterns, motivations, and implications of AI use among UoG undergraduates through four core questions: how students integrate AI into learning behaviors such as comprehension, assignment planning, exam preparation, and self-assessment; how they perceive AI’s influence on independent learning, critical thinking, motivation, and academic integrity; how usage varies by discipline, socio-economic background, language proficiency, and internet access; and how institutional or instructor-level factors encourage or constrain meaningfuluse. Using a mixed-methods convergent design, the study combined an online survey of 600 students with 30 semi-structured student interviews, 10 faculty interviews, and a review of institutional policies and digital-learning records. Quantitative data were analyzed through descriptive and regression techniques, while qualitative data were thematically coded. Results showed that nearly 70 percent of students used AI weekly, primarily for explaining concepts, brainstorming ideas, refining outlines, and simulating exam questions. STEM students valued AI for clarifying technical material and debugging, whereas Social-Science and Business majors used it for argument development and essay writing. Reported benefits included improved efficiency, confidence, and self-directed learning, but participants also warned of overreliance, shallow understanding, and blurred originality. Access disparities were pronounced: students with reliable connectivity and stronger English skills gained the most. Institutional factors also mattered; faculty who framed AI as a “study partner” fostered critical engagement, while unclear or absent policies generated confusion and misuse. Future research should track long-term cognitive and behavioral effects, experimentally compare AI-supported and traditional learning designs, explore inclusion for low-bandwidth and bilingual contexts, and evaluate institutional strategies such as prompt libraries, AI-integrated assessments, and faculty-training modules.