An Empirical Study on Policy Attitudes Toward Generative Artificial Intelligence in Teaching Assessment at Asian Universities Based on the Random Forest Algorithm

The application of generative artificial intelligence (GenAI) in teaching and assessment is becoming increasingly prevalent in Asian higher education institutions. However, significant differences exist among universities in terms of policy awareness and attitudes toward its use. This study analyzes 354 survey responses collected from universities in China, Japan, South Korea, India, and Pakistan, employing the random forest algorithm to examine respondents’ attitudes toward the future of AI and the key influencing factors. The results show that 75.14% of respondents hold a positive attitude. The model achieved an F1 score of 0.84, indicating strong predictive performance. Feature importance analysis reveals that perceived learning efficiency, awareness of content-type differentiation, and willingness to participate in policy-making are the main drivers of attitude formation. The findings provide empirical evidence to support AI governance and policy development in higher education.