Abstract Vocabulary acquisition plays a pivotal role in foreign language proficiency. This study investigates the effectiveness of Artificial Intelligence (AI)-powered mind mapping on English vocabulary acquisition among non-English-major undergraduates at Nguyen Tat Thanh University (NTTU). By employing a quasi-experimental mixed-methods design, 20 third-year students were assigned to an Experimental Group (EG), which utilized AI-powered mind-mapping tool (GitMind), and a Control Group (CG), which followed traditional vocabulary instruction methods. Quantitative data from pre-tests and post-tests indicated that the EG achieved significantly higher vocabulary gains than the CG (p = .044), with a medium-to-large effect size (Cohen’s d = 0.66). Qualitative findings derived from Technology Acceptance Model-based questionnaires and focus group interviews revealed high perceived usefulness, enhanced learner confidence, and increased engagement. The findings suggest that AI-powered visual mapping can reduce extraneous cognitive load and facilitate deeper semantic processing, thereby supporting vocabulary acquisition in EFL higher education contexts.