WHEN ALGORITHMS ADVISE: THE ROLE OF ALGORITHMIC TRUST AND PERCEIVED TRANSPARENCY IN SHAPING INVESTORS' RELIANCE ON AI-GENERATED FINANCIAL ADVICE

Abstract

While artificial intelligence is increasingly creating financial advice via robo-advisors, algorithmic tools and generative AI assistants, investors are not always taking the advice. This study investigates the influence of perceived AI transparency on investors' trust in AI-driven financial advice, as well as the role of algorithmic trust in this relationship. The cross-sectional survey was given to 310 eligible active retail investors who were introduced to AI-based financial advice tools online in investment communities and at investment-club events, based on trust-formation theory in HAI. Confirmatory factor analysis, reliability and validity tests, and bootstrapped mediation regression (5,000 resamples) were used to analyse data. Perceived AI transparency was found to have a modest but significant relationship with algorithmic trust (β = .165, p = .004), which in turn had a significant relationship with reliance on AI-generated financial advice (β = .376, p < .001). The effect of transparency on reliance was also direct, and was similar in magnitude to the effect mediated by trust (β = .133, p = .012). The transparency–reliance relationship was partially mediated by algorithmic trust (indirect β = .062, 95% CI [.021, .114]), which accounted for about 32% of the total effect of transparency, and the direct effect of transparency on trust and reliance was comparable in strength. The results indicate that transparency influences reliance in two parallel ways, which has implications for explainable AI design, financial institutions' AI deployment, and the regulation of transparency in FinTech.