Aligning stakeholders in AI-enabled customer service: Toward human-centric adoption in electronic marketplaces

Abstract Artificial intelligence (AI) is increasingly integrated into electronic marketplaces and digital platforms, where customer service functions as a critical interaction layer mediating trust, value creation, and user retention. While the strategic potential of AI is well recognized, most adoption models focus narrowly on organizational or user-centric perspectives, overlooking the stakeholder frictions that arise in real-world service environments. To address this limitation, we examine AI adoption in customer service through a multi-actor lens that captures the views of end users, frontline agents, and supervisors. Grounded in stakeholder theory and classical technology adoption models, the study employs a multi-stakeholder, exploratory design, combining quantitative online surveys ( n = 198) with illustrative qualitative insights from semi-structured interviews. Results reveal persistent misalignments across stakeholder groups. Users support AI when paired with human fallback, emphasizing empathy and trust. Agents express skepticism due to increased workload, limited training, and unclear communication. Supervisors prioritize efficiency, often underestimating operational frictions. These tensions are particularly relevant in AI-enabled marketplaces, where customer service operates as a key interface between platforms and users rather than as a standalone organizational function. Our findings reframe AI adoption as an organizational transformation that requires aligning technological efficiency with human experience. Rather than treating implementation as a purely technical upgrade, we highlight the need for human-centric strategies that balance automation with empathy and workforce inclusion. The study offers actionable insights for platform operators seeking sustainable, trust-based AI integration.