Faculty Perceptions and Practical Challenges in Integrating Generative AI into Business Education

Abstract

This study examines faculty perceptions and practical challenges regarding generative AI in business higher education. It employs a mixed-methods analytical paradigm that combines a qualitative research design with natural language processing (NLP) techniques to conduct a systematic analysis of in-depth, semi-structured interview data from 10 business faculty members. The study comprehensively employed word frequency statistics, the LDA topic model, and SnowNLP sentiment polarity analysis to characterize, from multiple dimensions, the faculty group's cognitive landscape, usage logic, and practical dilemmas regarding generative AI tools. The results show that faculty members generally hold a highly positive attitude toward AI (with a mean sentiment score of approximately 0.98); however, at the practical level, they exhibit an implicit usage pattern characterized by openness in efficiency-related scenarios and convergence in evaluation-related scenarios. There is a significant stratification in the dimensions of focus among faculty members of different ranks: lecturers emphasize operational skills and tool integration, while associate professors place greater emphasis on content quality control and ethical standards; LDA thematic analysis identified three core issues: the efficiency of case-based teaching, the cultivation of business simulation and analytical skills, and training needs and assessment reform, collectively revealing the inherent tension between efficiency improvement and competency development. This study supplements existing qualitative evidence on AI-enabled education research from the perspective of faculty members and provides empirical references for business schools to formulate differentiated training strategies and optimize teaching reform plans for human-machine collaboration.