Unveiling the Impact of AI-Chatbot Attributes and Anthropomorphic Cues in E-Commerce

Abstract

The rise of AI chatbots in e-commerce has revolutionized digital customer service, yet understanding the factors influencing customer adoption remains a challenge. Drawing on the Technology Acceptance Model 2 and Anthropomorphism theory, the study aims to examine how human-like interaction elements can enhance trust and usability, leading to greater customer engagement with AI chatbots in e-commerce. A quantitative approach was employed, involving a sample of 423 respondents from Malaysia who interacted with AI chatbots on e-commerce platforms. Data were analyzed using partial least square structural equation modeling (PLS-SEM). Findings revealed that perceived usefulness (PU) and perceived ease of use (PEOU) have a significant direct influence on customer adoption, where PU became the strongest predictor, followed by PEOU. Perceived trust showed no direct influence, but strongly predicted anthropomorphic cues which in turn had a direct positive influence on customer adoption and indirect positive influence as a mediator between trust and adoption. The variable perceived trust (PT) showed no direct influence on adoption; instead, anthropomorphic cues and a strong mediating effect by anthropomorphic features such as emotional expressions and human-like communication enhance user experience, build trust, and drive adoption. This research highlights the originality of the study, which clearly explains that functional attributes (PU and PEOU) are not solely determined by adoption but are carefully designed by the mediating variable of anthropomorphic features. These insights are critical for e-commerce businesses and AI developers in designing chatbots that foster customer trust, usability, and satisfaction. The study contributes to the academic discourse on AI adoption and offers practical strategies for optimizing chatbot-human interaction in digital commerce.