Negotiation with AI: Neutrality or Reproduction of Gender Roles?

Abstract: Drawing on Social Role Theory, this study examines whether AI-mediated salary negotiations disrupt gendered role expectations and reduce disparities in negotiation outcomes. In a simulation negotiating with an AI-driven chatbot, we analyzed gendered language patterns, negotiation performance, and perceptions of the AI counterpart. Findings revealed no significant gender differences in negotiation outcomes; however, communication style, particularly agentic and competitive language, correlated more strongly with outcomes, suggesting that traditional role-congruent behaviors still matter even in AI contexts. Participants frequently projected gender onto the AI, often aligning it with their own identity, and those perceiving the AI as male reported higher satisfaction. These results extend Social Role Theory into human–AI interaction by showing that while AI can reduce interpersonal pressures, implicit biases favoring masculine-coded negotiation norms persist. The study underscores the need for careful AI design and oversight to prevent reproducing gender inequality in emerging negotiation environments.

Keywords: gender gap, artificial intelligence (AI), communication style, stereotype threat, human-AI interaction