AbstractAs predictive artificial intelligence (AI) technologies increasingly steer workplace decisions, debates around fairness have intensified. Existing research often approaches fairness either as a set of universal principles supported or undermined by algorithms, or as a product of social interpretations, thereby providing either technologically deterministic or purely social accounts. Drawing on an ethnographic study of a human resources (HR) department of a large international company that introduced AI in hiring, this study offers an alternative view that shifts focus to how fairness emerges through the ways people define, embed, and perform values with algorithms. Taking a sociomaterial perspective, we find that the introduction and use of AI resulted in crowding out expert practices of performing fairness, favouring instead the version performed by HR. Our process model explains this outcome by the growing symbiosis between HR's professional mandate for fairness and AI procedures, where each legitimizes, shapes, and protects the other over time. This study thus shows that fairness is not pre‐given but constantly redefined and enacted through evolving associations between professional mandates and AI technologies.

This paper presents research on how knowledge brokers attempt to translate opaque algorithmic predictions. The research is based on a 31-month ethnographic study of the implementation of a learning algorithm by the Dutch police to predict the occurrence of crime incidents and offers one of the first empirical accounts of algorithmic brokers. We studied a group of intelligence officers, who were tasked with brokering between a machine learning community and a user community by translating the outcomes of the learning algorithm to police management. We found that, as knowledge brokers, they performed different translation practices over time and enacted increasingly influential brokerage roles, namely, those of messenger, interpreter, and curator. Triggered by an impassable knowledge boundary yielded by the black-boxed machine learning, the brokers eventually acted like “kings in the land of the blind” and substituted the algorithmic predictions with their own judgments. By emphasizing the dynamic and influential nature of algorithmic brokerage work, we contribute to the literature on knowledge brokerage and translation in the age of learning algorithms.

Because new technologies allow new performances, mediations, representations, and information flows, they are often associated with changes in how coordination is achieved. Current coordination research emphasizes its situated and emergent nature, but seldom accounts for the role of embodied action. Building on a 25-month field study of the da Vinci robot, an endoscopic system for minimally invasive surgery, we bring to the fore the role of the body in how coordination was reconfigured in response to a change in technological mediation. Using the robot, surgeons experienced both an augmentation and a reduction of what they can do with their bodies in terms of haptic, visual, and auditory perception and manipulative dexterity. These bodily augmentations and reductions affected joint task performance and led to coordinative adaptations (e.g., spatial relocating, redistributing tasks, accommodating novel perceptual dependencies, and mounting novel responses) that, over time, resulted in reconfiguration of roles, including expanded occupational knowledge, emergence of new specializations, and shifts in status and boundaries. By emphasizing the importance of the body in coordination, this paper suggests that an embodiment perspective is important for explaining how and why coordination evolves following the introduction of a new technology.

When actors deem technological change undesirable, they may act symbolically by pretending to comply while avoiding real change. In our study of the introduction of an algorithmic technology in a sales organization, we found that such symbolic conformity led unintendedly to the full implementation of the suggested technological change. To explain this surprising outcome, we advance a regime-of-knowing lens that helps to analyze deep challenges happening under the surface during the process of technology introduction. A regime of knowing guides what is worth knowing, what actions matter to acquire this knowledge, and who has the authority to make decisions around those issues. We found that both the technologists who introduced the algorithmic technology, and the incumbent workers whose work was affected by the change, used symbolic actions to either defend the established regime of knowing or to advocate a radical change. Although the incumbent workers enacted symbolic conformity by pretending to comply with suggested changes, the technologists performed symbolic advocacy by presenting a positive side of the technological change. Ironically, because the symbolic conformity enabled and was reinforced by symbolic advocacy, reinforcing cycles of symbolic actions yielded a radical change in the sales' regime of knowing: from one focused on a deep understanding of customers via personal contact and strong relationships, to one based on model predictions from the processing of large datasets. We discuss the theoretical implications of these findings for the introduction of technology at work and for knowing in the workplace.