This study explores the application of AI models in grading bachelor’s and master’s level assessments, comparing their performance with traditional human grading methods. The research aims to evaluate the strengths and limitations of AI-driven grading systems, focusing on key metrics such as accuracy, efficiency, reliability, and fairness.

Through a quantitative experimental approach, this study analyzes primary data from academic assessments to determine AI's ability to handle complex, nuanced academic content. Furthermore, it investigates challenges such as biases in AI training data and the systems' capacity to evaluate critical thinking and creativity. By addressing these issues, the research aims to provide actionable insights for integrating AI into higher education, contributing to a more efficient and equitable academic evaluation process.

The findings are expected to guide future studies and support the development of improved AI-based grading technologies.