ABSTRACT
The adoption of Generative Artificial Intelligence (GenAI) is reshaping the very nature of work, yet its social meaning and evaluative implications—particularly for women—remain underexplored. Although existing research has mainly examined GenAI adoption, use, and algorithmic bias, less is known about how the use of GenAI shapes how individuals are evaluated by others. Drawing on the Stereotype Content Model, Role Congruity Theory and techno-feminism perspectives, this study examines how women’s use of GenAI influences perceptions of competence and warmth in academia. Using a 2 × 2 × 2 experimental vignette design [gender*GenAI adoption*task type (teaching vs. research)] with 651 participants, we examine how evaluations vary by academic gender, GenAI adoption, and task domain (teaching vs. research). We find that GenAI adoption generally enhances perceived competence. Women, however, gain the most in research tasks, where technical skills challenge gendered expectations. Conversely, GenAI use reduces perceived warmth, disproportionately penalizing women, especially in research. These findings reveal a “double bind,” highlighting how GenAI adoption may reproduce and intensify gendered evaluative tradeoffs within academic prestige hierarchies and the need for institutional strategies to mitigate relational penalties while recognizing competence.