He or she? A male-default bias in Chatbot gender attribution across explicit and implicit measures

As chatbots become increasingly integral to human–computer interactions, questions arise about how gender stereotypes shape users' perceptions. Across two studies, this research investigated whether users exhibit gendered biases toward chatbots and how these biases interact with occupational contexts. In Study 1 ( N = 66), participants evaluated visually embodied chatbots embedded in gender-typed roles, rating each chatbot's competence and their overall attitude toward it. Study 2 ( N = 91) utilized a mouse-tracking paradigm to assess implicit gender attributions toward a gender-neutral chatbot. Findings revealed a robust male-default bias: participants were more likely to perceive chatbots as male—in the absence of gender cues—and this bias persisted across contexts. Biases emerged in both explicit evaluations and implicit behavior, suggesting that gender stereotypes are activated automatically yet remain consciously accessible. These findings show how societal gender schemas extend to non-human agents and highlight the contextual triggers of gendered cognition and underscore the importance of considering implicit gender biases in the design of chatbot interfaces to promote more inclusive human–AI interactions. Highlights • Users consistently exhibited a male-default bias when evaluating gender-neutral chatbots. • Occupational context moderately influenced gender perception, but the male-default bias still persisted across conditions. • The results of implicit mouse-tracking measures aligned with explicit responses, revealing strong stereotype accessibility. • Findings reveal the extension of gender schemas to non-human agents, informing gender-fair chatbot design.