Augmented analytics shifts much of the analytical work in business intelligence (BI) to artificial intelligence (AI), leaving non-technical users with one remaining safeguard: verifying AI-generated insights before acting on them. This paper examines whether trust in augmented analytics is associated with less verification and how verification relates to perceived decision quality. Survey data from 350 non-technical BI users in Vietnam were analysed with partial least squares structural equation modelling (PLS-SEM), using a newly developed ten-item verification behaviour scale (VER). The results suggest that trust is negatively associated with verification, and verification is positively associated with perceived decision quality. The indirect association of trust through verification is negative, whereas its direct association remains positive, consistent with competitive mediation. Perceived explainability is associated with higher trust, but its direct association with verification is not significant, and the hypothesized moderating role of AI literacy receives no statistical support. Because the design is cross-sectional and self-reported, the paths represent associations rather than causal effects. Automation complacency and objective decision quality were not measured, and the verification scale was supported by initial validation evidence from one sample. The findings motivate linking insights to their source data, second checks for high-value decisions and direct monitoring of verification; the effectiveness of these proposals remains untested. The study develops a survey measure of verification behaviour with initial validation evidence and shows that the positive trust–decision quality association coexists with a negative indirect association through verification.