AI-chatbots as frontline agents promise innovative opportunities for shaping service offerings that benefit customers and retailers. Examining current practice through the lens of agency, as defined by Social Cognitive Theory, we present a 3-level classification of AI-chatbot design (anthropomorphic role, appearance and interactivity) and examine how the combination of these three aspects of chatbot design impacts on the complementarities of agency. Recognizing current implementation challenges, we advance that the complementarities of agency at each level are the lynchpin mechanism that translates AI-chatbot design into service relevant outcomes. We develop a research agenda focused on the emotion interface, resolution of the proxy agency dilemma and development of collective agency to support the implementation of AI-chatbots as frontline service agents.

作为一线代理的人工智能聊天机器人承诺提供创新机会，以塑造有利于客户和零售商的服务产品。通过社会认知理论定义的代理视角检查当前的实践，我们提出了人工智能聊天机器人设计的 3 级分类（拟人化角色、外观和交互性），并研究了聊天机器人设计的这三个方面的组合如何影响代理的互补性。认识到当前的实施挑战，我们提出，各级机构的互补性是将 AI 聊天机器人设计转化为服务相关结果的关键机制。我们制定了一个专注于情感界面的研究议程，