Navigating the "Automation Frontier": Relational Work and AI Orientations among Advanced College Students

The emergence of generative artificial intelligence (AI) technologies has sparked debates about how tasks should be divided between humans and AI. This paper examines how advanced college students navigate human-AI boundaries in the present and in imagined career futures by analyzing the “relational work” (Christin and Petre 2020; Zelizer 2012) they undertake to align their perceptions of AI, activities, and AI use across academic and pre-professional contexts. Drawing on Zelizer’s relational work framework to analyze semi-structured interviews with 28 upper-division, graduate, and professional students, I identify three predominant orientations toward AI: “Resisters” who broadly reject AI on moral grounds, “embracers,” who perceive AI as an efficiency-enhancing collaborator, and those who make “viable matches” by negotiating the moral conflicts they associate with AI through selective adoption. These orientations are shaped by institutional legitimation and guidance, enabling experiences with AI, moral concerns, and desires to protect valued skills for anticipated futures.