Agentic ai attributes in hedonic, heuristic, utilitarian and systematic processes: the role of trust in shaping user acceptance of agentic ai robo-advisory

Lee, Khai Xin (2025) Agentic ai attributes in hedonic, heuristic, utilitarian and systematic processes: the role of trust in shaping user acceptance of agentic ai robo-advisory. Final Year Project, UTAR.

PDF Download (1241Kb)

Abstract

The rapid rise of Agentic Artificial Intelligence (AI) has transformed financial technology by enabling robo-advisory systems that provide autonomous, adaptive investment guidance. Despite these advances, user acceptance remains contingent on trust, particularly within high-stakes financial investment decision-making. This study investigates how agentic AI attributes, expressed through hedonic, utilitarian, and systematic processing, shape affective and cognitive trust in robo-advisors, leading to user acceptance. Furthermore, this study examines respondents’ risk tolerance to uncover behavioral patterns that shape the acceptance of agentic AI robo-advisors. Guided by the Heuristic–Systematic Model (HSM), this study adopts a positivist philosophy and employs a deductive and quantitative approach. The rapid rise of Agentic Artificial Intelligence (AI) has changed financial technology by allowing robo-advisory systems to provide automated and adaptive investment guidance. However, user acceptance still depends on trust because financial investment decisions involve high risk. This study examines how agentic AI attributes, shown through hedonic, heuristic, utilitarian, and systematic processing, shape affective and cognitive trust in robo-advisors and lead to user acceptance. In addition, this study looks at respondents’ risk tolerance to identify behavior patterns that influence the acceptance of agentic AI robo-advisors. This study is based on the Heuristic–Systematic Model (HSM) and uses a positivist philosophy with a deductive and quantitative approach. Data collected via a cross-sectional survey targeted Generation Z (ages 18–28) in Malaysia. With 273 responses obtained and analysed using Partial Least Squares Structural Equation Modelling (PLS-SEM), the results confirm that Perceived Pleasure and Anthropomorphism have a significant impact on effective trust. Meanwhile, except for functional, Informational, Perceived Knowledge, and Perceived Benefit were found have a significant impact on Cognitive Trust. It also confirmed that affective and cognitive trust lead to acceptance of robo-advisory use. This research contributes theoretically by clarifying the dual roles of affective and cognitive trust in shaping the adoption of Agentic AI robo-advisors, and practically by guiding fintech firms and policymakers in designing transparent, user centred financial AI systems.

Actions (login required)