AI-Driven Service Quality in E-Commerce: Expectation–Perception Gaps Using Interval-Valued Pythagorean Fuzzy Analysis

Digital commerce is transforming through Artificial Intelligence (AI), which has redesigned the interaction with customers, delivery of services, and personalizing of customer experience. Nevertheless, the expectations and perceptions of the customers regarding the quality of the AI-driven service (AISQ) are still poorly comprehended, in particular, in regard to trust and reliability throughout the digital purchasing experience. This paper analyses consumer expectations and consumer perception of AISQ in e-commerce through Interval- Valued Pythagorean Fuzzy (IVPF) framework to deal with uncertainty in human judgments. According to 684 online customers, the results of this study reveal that there are wide differences between the expected and perceived performance of AI and that the highest degree of dissatisfaction is registered in terms of efficiency of problem solving, personalization with contexts, timely updates and emotional assurance with the uniformity of cross-platform system, which is satisfactory. The research has both theoretical and methodological implications in that it analyses the AISQ dimensions gaps as a multidimensional construct, and methodologically by enhancing gap analysis by IVPF to include trust and user satisfaction assessment in online commerce. In practice, the results give practical recommendations to the improvement of reliability, responsiveness, personalization, privacy assurance, and empathetic interaction design, allowing e-commerce companies to provide more reliable and user-friendly AI-based services to people in various regional markets.