This research endeavours to elucidate the intricate interplay between the perceived realism of deepfake simulations, engagement with learning materials, and financial decision-making confidence through a sophisticated methodological framework employing Structural Equation Modeling (SEM) and bootstrapping techniques. By leveraging a rigorously validated measurement model, the study dissects the latent constructs underpinning cognitive assimilation within financial pedagogy, positing that the verisimilitude of digital simulations engenders heightened epistemic engagement, subsequently fortifying individuals’ confidence in fiscal decision-making. The empirical architecture is anchored in multivariate statistical paradigms, ensuring robust parameter estimation and construct reliability. The findings substantiate the proposition that perceived realism exerts a direct influence on both engagement and confidence, with engagement serving as a pivotal mediating conduit. Covariance analysis delineates interdependencies among constructs, reinforcing the inferential robustness of the hypothesised relationships. Factorial loadings and construct reliability indices affirm the internal consistency and discriminant validity of the employed psychometric instruments. Bootstrapped path coefficients further corroborate the statistical significance of the hypothesised linkages, fortifying the generalizability of insights. This study contributes to the epistemological corpus by bridging the domains of cognitive perception, digital realism, and financial acumen, offering managerial implications for the design of immersive learning interventions that bolster informed financial behaviours.