Embedding AI into human resource training and development practices in Taiwan

Integrating AI into employee training transforms HRM, yet existing frameworks overlook AI-specific metrics and employee well-being. This study investigates AI-driven training in ten Taiwanese firms, a context marked by extreme IP protection, high error costs, collectivist culture, and talent competition. We develop an empirically grounded framework integrating engagement traces, learning trajectories, and sentiment analytics into a hybrid evaluation model. Four transformation mechanisms are identified: hyper-personalization, predictive risk-to-intervention translation, tacit knowledge codification, and real-time feedback. Findings reveal that AI significantly compresses time-to-competency and enhances resilience but generates an autonomy paradox requiring transparent governance and cultural adaptation.