Member Only Access
Summary
As generative AI tools become embedded in advertising and marketing research workflows, questions about bias increasingly extend beyond outputs to the interaction itself. This study examines whether gendered patterns can enter AI through subtle differences in how prompts are phrased. By systematically varying linguistic styles using psychologically grounded traits, the research shows that implicit, style-based, gender cues shape AI prompt construction more strongly than explicit, gender labels, with important implications for how bias may propagate upstream in AI-assisted marketing and research applications.
