Veranstaltungsprogramm der VHB Jahrestagung

AI-supported hiring is regularly lauded as a way to eliminate human bias in the hiring process. Such a stance is aligned with techno-optimism where technology is seen as a force for good. While some see AI-supported hiring as a way to reduce human bias, AI-supported hiring is commonly singled out as an example of algorithmic bias, which creates and amplifies inequalities. This article traces the question of how techno-optimism in relation to AI-enabled hiring can be maintained in spite of the fact that algorithmic bias exists. The study draws on interviews with those who design and use such technologies to shed light on how algorithmic bias is discussed in relation to AI-supported hiring. The article shows that techno-optimism is maintained by stressing that humans make hiring processes biased and by constructing algorithmic bias as fixable through a range of techniques such as fixing the data, fixing the human, ‘blinding’ the algorithm and various processes of quality control and auditing algorithms. By constructing algorithmic bias as something that is solvable it is possible to maintain a techno-optimistic stance. Furthermore the article shows that rather than adopting techno-pessimism, the alternative stance to techno-optimism is techno-hesitation. Techno-hesitation stresses that the collaboration between humans and AI leads to improved human decision making but that short-term reputational concerns impede the use of AI-supported hiring. Techno-hesitation is a temporal stance in that it is expected that once AI in hiring becomes commonplace and algorithmic bias is fixed, reputational concerns will dissipate. The paper shows that such an understanding conceives algorithmic bias as a technical problem for which a technical solution is sought rather than seeing technology as shaped by and shaping society, which would allow for a more nuanced and complex conceptualisation of algorithmic bias to emerge.

New technological developments are leading to the increased use of robots in the service sector. Yet, literature about service employees’ reactions to robot introductions is still scant. This study builds on and expands this literature based on a field experiment with the humanoid robot Pepper deployed in a service setting for six weeks. It presents an exploratory case study of a service company with multiple sites, where a humanoid service robot was implemented in only one of the sites. Data from semi-structured interviews is analyzed using a grounded theory approach. It is complemented by comparing employees’ affective and cognitive reactions to the implementation between those who experienced the implementation on their own and those who did not. In general, employees who interacted with Pepper tended to have a more positive affective and cognitive reaction than those who had no interaction. Additionally, they advocated for the permanent deployment of the robot to a larger extent. This work contributes to a better understanding of the possible consequences of service robot implementation for employees and how management may navigate through times of change and transformation caused by increasingly intelligent technologies. Further, it expands current literature by looking at the organization and its workforce in their entirety rather than just those directly affected by the implementation. Therefore, this study reveals different, employee reactions to the implementation of the very same technology, depending on whether employees have directly experienced the focal technology, or have only indirect information through other channels (e.g., hearsay).