From Assistance to Accountability: Generative AI in Service Work

This project examines how customer support staff use generative AI tools that suggest responses during live service interactions, and how responsibility for what is ultimately said remains with human workers. Focusing on customer support operations in Japan, the study looks at how agents decide whether to accept, modify, or reject AI-generated suggestions when handling real customer concerns, especially in interactions that are recorded and later reviewed for quality, compliance, or training. By closely observing everyday service work, including chat logs, call recordings, and on-screen use of AI tools, the project will clarify when and how AI assistance actually improves service performance, supports professional judgment, and protects workers under demanding conditions. Participating institutions are expected to benefit from concrete insights into how AI tools are used in practice, which can inform training, quality assurance, and responsible deployment strategies that improve service consistency while maintaining accountability and trust with customers.

Faculty Supervisor:

Parbudyal Singh

Student:

Partner:

Kyoto University

Discipline:

Sociology

Sector:

Education

University:

York University

Program: