Students are increasingly using Generative Artificial Intelligence (GenAI) for learning, problem solving and understanding statistical concepts. However, the effective use of such tools requires the ability to critically assess and validate AI-generated solutions. In this study we investigate GenAI use among undergraduate Statistics students, focusing on Perceived Learning Benefits (PLB), Verification Behavior Score (VBS) and confidence in judging AI-generated statistical solutions. Data were acquired from 294 undergraduate students (244 GenAI users) through a structured online questionnaire. Perceived learning advantages and verification behavior were assessed using multi-item measures. A Random Forest classification model was constructed to predict students’ confidence in judging AI-generated statistical answers from AI-use patterns, learning advantages, verification behavior, Statistics experience and year of study. We tested the model with a stratified 5-fold cross-validation. The results point to the pedagogical value of GenAI and the need for verification and critical evaluation skills in AI-assisted learning of Statistics.