Gendered Cognitive Processing of Sexual Appeal in Advertising: AI-Generated Virtual Humans vs. Real Human Models – A Multimodal fNIRS and Eye-Tracking Study

Despite the growing use of digital agents in marketing, empirical comparisons between consumer responses to AI-generated Virtual Humans and Real Human Models remain limited. This study investigated gender-specific processing of sexual appeal advertisements using surveys, functional near-infrared spectroscopy (fNIRS), and eye-tracking in 33 university students. Male participants rated AI-generated Virtual Humans as highly attractive while exhibiting increased dorsolateral prefrontal cortex activation, reflecting higher cognitive regulation costs, a pattern termed Cognitive-Evaluative Dissociation. In contrast, female participants demonstrated a Psychophysiological Integration pattern, in which heightened physiological arousal, indexed by pupil dilation, was negatively associated with subjective evaluations. These findings indicate that males experience a divergence between explicit preference and implicit cognitive load, whereas females rely more on intuitive safety-based assessments of virtual entities. Overall, the results underscore the importance of gender-sensitive strategies in AI-based advertising, emphasizing cognitive load reduction for males and narrative contextualization for females.