Humanizing Artificial Intelligence: Exploring Explainability, Attribution, and Sustained Trust in Agentic AI Systems

Abstract

Agentic AI systems are reshaping the fintech advisory landscape by combining autonomy, adaptability, and goal-directed intelligence with relational sensitivity. Unlike conventional tools, Agentic AI represents a story of three-layered evolution: technological, human, and institutional— each enriching the other. Technologically, it advances explainability, adaptability, and interaction; humanly, it fosters relational trust, empathy, and confidence; institutionally, it supports the development of adaptive governance, responsible regulation, and sustainable ethical frameworks that enable long-term trust and innovation.This study constructs a model to examine the impact of social interaction cues and anthropomorphic factors on users’ sustained trust by integrating the Computers As Social Actors (CASA) theory with attribution theory. An empirical analysis of 262 survey responses reveals that CASA factors— anthropomorphic characteristics, relational experience and system design factors like system transparency, system reliability & performance, and user experience, collectively enhance attribution, which in turn sustains trust and encourages long-term engagement with AI services. Agentic-AI systems function as mediators, translating these relational and system-level cues into attributional judgments that strengthen user trust and loyalty. The findings expand the scope of human–AI interaction research, offering insights for fintech and customer service ecosystems where trust, transparency, and relational engagement are pivotal.