The Relationship Between Performance and Trust in AI in E-Finance

No prior work has fully resolved how individual performance metrics shape user reliance on automated financial systems. It was already known that technological complexity often hinders seamless human-machine collaboration. Prior research has shown that trust remains a primary barrier to widespread adoption of digital tools. That uncertainty drove this investigation into the specific dynamics of robo-advisor interactions. This gap motivated a closer look at how users weigh their own skills against machine output. Prior studies frequently overlooked the relative performance difference between human and machine agents. Most existing literature focuses on singular performance indicators rather than comparative outcomes. This study addresses these limitations by examining the interplay between user competence and algorithmic efficacy.

This study aims to analyze the effect of performance on trust in a robo-advisor through an empirical investment simulation. The researchers sought to clarify how users perceive and rely on automated financial tools. They investigated whether individual performance metrics or comparative outcomes better explain trust dynamics. The motivation stemmed from the increasing role of artificial intelligence in personal finance management. The team addressed the complexity of human-machine collaboration in high-stakes environments. They aimed to determine if social biases influence how individuals interact with non-human agents. The study sought to identify the conditions under which users trust automated systems. This work provides insight into the psychological factors shaping digital financial adoption.

The research team implemented an empirical investment simulation to evaluate user behavior. This experimental design allowed for the systematic manipulation of performance variables. Participants engaged with an automated system to manage financial portfolios under varying conditions. The investigators recorded trust levels at multiple intervals throughout the simulation. They analyzed the data to determine how relative success influenced user reliance. The approach focused on comparing human outcomes against machine-generated results. This methodology enabled the isolation of performance gaps as a primary variable. The study design ensured that participants faced realistic financial decision-making challenges.

The relative performance gap between humans and machines acts as a moderate indicator of trust changes. Individual performance metrics for either the human or the AI serve as weak predictors. The data shows that users weigh their own success against the machine's output. When capabilities are comparable, the difference in outcomes drives trust fluctuations. The researchers observed that low transparency facilitates the emergence of social biases. These biases mirror those found in human-to-human interactions. The findings indicate that users do not rely on absolute performance alone. Instead, the comparative delta between the two agents defines the user experience.

The authors propose that relative performance serves as a moderate predictor of trust when human and machine capabilities align. They suggest that individual performance metrics alone provide limited insight into user confidence levels. The researchers observe that low transparency in algorithmic decision-making may trigger social biases. These biases appear to mirror patterns typically observed in human-to-human social dynamics. The study implies that transparency is a key factor in mitigating biased perceptions of automated systems. They conclude that users do not simply evaluate AI based on absolute success rates. Instead, the comparative gap between the user and the system influences their willingness to trust. These findings provide a framework for understanding human-AI collaboration in financial contexts.