Perceived AI Service Attributes, Consumer Trust, and AI-Enabled Financial Decision Support in Chinese Financial Services

Abstract

Trust becomes a crucial factor in financial decision support systems, as AI-driven financial services become more prevalent in delivering information and evaluating choices, while also providing automated guidance. In this exploratory study, the relationship between perceived AI service attributes, psychological evaluations and institutional factors, as well as consumer trust in China were analyzed by analyzing 402 questionnaires obtained from the retained population. Analyses consisted of descriptive statistics, measurement diagnostics, correlations, and three sequential ordinary least squares regression models with HC3 heteroskedasticity-robust inference. In the model with just attributes, competence, transparency, and security were positively related to consumer trust, accounting for 51.45% of the variance observed. Once these factors were accounted for, security was still significantly related to trust, but competence and transparency were not. The final model accounted for 61.08% of the variability observed, with usefulness, ease of use, perceived risk, and institutional trust being positively associated with the model. There was no significant difference overall between duration and frequency. The overall results indicate that perceived protection, institutional assurances, usability, and psychological assessments are all relevant to the level of trust in financial decision support enabled by AI. Consumer trust is not viewed as evidence of acceptance of consumer recommendations, actual financial choice, reliance behaviour, or decision quality but rather as an attitudinal outcome relevant to human – AI reliance.