As organizations integrate AI into research, synthesis, blueprinting, and specification writing, the promise is speed, scale, and efficiency. Yet in service design, these activities do more than produce outputs: they are the very practices through which teams make sense of complexity. Interviews are not only data collection; they are relational encounters where trust is built, ambiguity surfaces, and unexpected signals emerge. Journey maps, blueprints, and specs are not merely deliverables; they are thinking tools that help teams connect evidence, expose assumptions, and develop shared judgment.

This talk asks: what happens when AI moves from supporting these practices to replacing them? Drawing on examples such as AI-led interviews, automated synthesis, and one-click blueprint generation, I argue that the central risk is not only inaccurate or generic outputs. It is the loss of the reflective and relational work through which service designers understand people, align organizations, and turn insight into meaningful change. Rather than rejecting intelligent technologies, I propose a more responsible role for AI in service design: augmenting human sensemaking, preserving human agency, and keeping lived experience at the center of how services are designed, delivered, and evolved.