Employment Negotiations With an Algorithm? How AI as Negotiation Counterpart Would Affect Negotiators' Trust and Subjective Value Expectations

Sonder, Dominik; Arnhold, Nadine; Hertel, Guido

Abstract

Artificial intelligence (AI) offers manifold ways to improve organizational recruiting and conflict resolution due to high processing power and low subjective biases. Additionally to AI applications approaching and selecting job applicants, AI usage might also be promising in employment negotiations as the final step of recruitment. However, it is largely unclear how applicants would react to the use of AI technology in such a context marked by high stakes and (at least partly) conflicting interests, which need to be resolved to terminate the recruitment process successfully. Therefore, we examined how AI negotiation agents influence negotiators' trust in and expectations about an upcoming employment negotiation. In a preregistered experimental vignette study, participants (n = 291) imagined to prepare for an employment negotiation. Varying the type of negotiation counterpart (human—AI with avatar—AI without avatar), we assessed participants' trust intentions and subjective value expectations regarding the anticipated negotiation. We found higher trust and more positive expectations for human counterparts as compared to AI counterparts, regardless of whether the AI was presented with or without an avatar. However, participants' technology-related attitudes and skills attenuated this effect, leading to less negative effects for AI as a negotiation counterpart when participants possessed high skills and positive attitudes regarding digital technologies. Overall, this study provides initial evidence of how people would react to AI agents in mixed-motive and conflict resolution settings and may serve as a starting point for developing evidence-based guidelines for designing AI agents for HR and conflict resolution settings.