Large Language Models in Higher Education: Evaluating Their Impact on Personalized Learning, Academic Performance, and Assessment Integrity

Original Research | 2026 | July | Volume 1 | Issue 2 | Page 50-62

Manoj Kumar

Email: mkumar75@gmail.com

Affiliation: Famotion Technologies Pvt Ltd

Designation: CVO

Large Language Models in Higher Education_ Evaluating Their Impact on Personalized Learnin.

Download • 266KB

Abstract

The rapid advancement of Large Language Models (LLMs), such as ChatGPT and other generative artificial intelligence (AI) systems, has significantly transformed teaching, learning, and assessment practices in higher education. These technologies provide personalized learning experiences by delivering adaptive explanations, real-time feedback, intelligent tutoring, and content generation tailored to individual student needs. However, their widespread adoption has also raised concerns regarding academic integrity, assessment validity, and the development of critical thinking skills. This study aims to evaluate the impact of LLMs on personalized learning, academic performance, and assessment integrity among higher education students. A cross-sectional mixed-methods study was designed involving undergraduate and postgraduate students from multiple academic disciplines. Quantitative data were collected through structured questionnaires measuring AI usage patterns, learning satisfaction, perceived academic improvement, and ethical concerns, while qualitative feedback explored students' experiences and faculty perspectives. Descriptive and inferential statistical analyses were performed to identify relationships between LLM utilization and educational outcomes. The findings indicate that regular use of LLMs significantly enhances personalized learning by improving conceptual understanding, study efficiency, and learner engagement. Students reported increased confidence in completing academic tasks and better access to learning support outside the classroom. Nevertheless, concerns emerged regarding excessive dependence on AI-generated content, reduced independent problem-solving, plagiarism, and challenges in maintaining assessment authenticity. Faculty participants emphasized the need for revised assessment strategies that prioritize critical thinking, creativity, and authentic performance-based evaluation. The study concludes that LLMs have considerable potential to enhance higher education when integrated responsibly within ethical and pedagogical frameworks. Institutions should establish comprehensive AI governance policies, promote AI literacy among educators and students, and redesign assessment methods to preserve academic integrity while maximizing the educational benefits of generative AI technologies.

Keywords: Artificial Intelligence, Large Language Models, Academic Performance, Assessment Integrity