The introduction of generative AI tools such as ChatGPT into creative workplaces has sparked highly visible, but binary worker replacement and augmentation debates. This study reframes this argument by examining how creative professionals re-specify a division of labor with these tools. Through 17 ethnomethodologically informed interviews with international creative agency workers we demonstrate how roles are assigned to generative AI tools, how their contributions are modified and remediated, and how workers practically manage their outputs to reflect assumptions of internal and external stakeholders. This paper makes 3 unique contributions to CSCW: (1) we conceptualize generative AI prompting as a type of workplace situated, reflexive delegation, (2) we demonstrate that workers must continuously configure and repair AI role boundaries to maintain workplace intelligibility and accountability; and (3) we introduce the notion of interpretive templatized trust, where workers devise strategies to adapt automated generative templates for their setting, and reinforce stakeholder trust. This contribution has implications for organizing productive human-AI work in creative and stakeholder centric environments.
Clarke et al. (Sat,) studied this question.