AI-Assisted Learning in Business Negotiation Education: A PLS-SEM Investigation Through Intelligence Augmentation Theory

Business negotiation courses have long struggled to bridge conceptual knowledge and enacted performance, and large language models now offer practice conditions that earlier technologies could not. This study examines how AI-assisted learning produces felt learning value in a vocational negotiation course in China, using Intelligence Augmentation Theory to frame two augmentation pathways: one proximal, through negotiation skill enhancement, and one distal, through innovative thinking development. A PLS-SEM model was tested with data from 69 students who used the Doubao AI assistant during case preparation. Three of four hypothesized paths were supported. Perceived AI application predicted negotiation skill enhancement strongly (β = 0.662), which in turn predicted perceived learning effectiveness (β = 0.445), with a significant indirect effect (β = 0.295). The path through innovative thinking development was weaker (β = 0.309) and did not carry through to learning effectiveness (β = 0.198, p = .313). Generative AI is widely positioned as a catalyst for higher-order thinking in education, yet direct evidence from professional skills courses remains scarce. The asymmetric results suggest that within a single-course window, AI-assisted learning produces felt value primarily through visible, assessable skill gains rather than creative development. This does not mean the distal pathway is inert, but that its payoff may operate on a longer time horizon than one course can capture. The findings offer an empirically grounded basis for calibrating expectations about what short-term AI integration can and cannot accomplish in vocational professional education.