The Fabricated Front: Generative AI and the Opacity of Workplace Performance

Workplace trust relies on communicative performances that act as diagnostic cues for competence, attention, and commitment. Drawing on Erving Goffman’s dramaturgical framework, professional communication functions as a "front" through which workers make these qualities available for interpretation. Generative AI (GenAI) disrupts this tacit infrastructure of coordination by introducing mediated communication that can modify or generate messages on behalf of workers. This paper extends the concept of effort opacity—the systematic decoupling of observable output from human engagement—by examining the interactional mechanics that produce opacity in everyday workplace encounters.

To specify where this decoupling occurs in interaction, the authors propose five opacity mechanisms divided into two distinct groups. Identity mechanisms concern who a worker appears to be and include voice (filtering stance cues and personality) and provenance (gaps in accountability and citation verification). Labor mechanisms index what a worker actually did and include vulnerability (filtering out hedges and pauses that invite mentorship), attention (performing attendance without cognitive engagement), and investment (making high- and low-effort outputs indistinguishable).

Analyzing 1,250 transcripts from Anthropic's AI Interviewer dataset across general workforce participants, creative professionals, and scientists, the authors reveal systematic variation in how different fields prioritize these mechanisms. Scientific writing heavily prioritizes provenance verification due to the risk of hallucinations, whereas creative fields focus on investment opacity regarding credit and craft. Across all groups, a profound identity/labor asymmetry emerges: professionals fiercely resist opacity around identity mechanisms like provenance and voice, yet overwhelmingly produce opacity around labor mechanisms to hide knowledge gaps and minimize visible effort.