AI and fairness: An ethnographic study on AI-based hiring

The increasing use of artificial intelligence (AI) in organizations has sparked debates about its potential to automate and augment knowledge work. However, less attention has been paid to how AI is deployed insituations where moral issues of fairness are at stake.

This studybuilds on a three-year ethnography of a large international company's
human resources (HR) department that introduced an AI-based tool to improve fairness in the hiring process. We found that the use of AI for automating selection decisions created tensions between how the AI tool encoded fairness and how hiring managers understood fairness in practice. Surprisingly, the HR department responded to these tensions not by moving to augmentation, but by retaining automation and instead
redesigning the technology.

Our study, therefore, offers an alternative scenario to tensions arising from AI automation, where the organization decides to make automation more human-centered. By opening the black box surrounding the debate around AI technologies and fairness, this study contributes to the management literature on AI automation and augmentation and studies on morality and emerging technologies.