Presentation
GAN-Generated Synthetic Faces May Induce Implicit Processing
SessionPoster Session 2
DescriptionArtificial intelligence (AI)-generated synthetic images, particularly those produced by Generative Adversarial Networks (GANs), are increasingly common across social platforms and human–AI interfaces. These images present both opportunities and challenges, including cybersecurity risks, misinformation, and issues of trust and accessibility. The present study examined how affective processing and emotional competence influence the perception of GAN-generated face images. Using an affective face perception task, self-report measures, and an eye-tracking task, we investigated trust ratings and gaze fixation patterns when participants viewed real and synthetic faces. Our findings partially supported our hypotheses. While trust ratings did not differ significantly between real and synthetic faces, participants spent longer viewing synthetic faces, suggesting that implicit processing contributes to their perception. Higher affective face perception accuracy predicted greater trust in both real and synthetic faces, while emotional competence was associated with increased trust in synthetic faces relative to real ones. These findings have important implications for human–AI interaction and cybersecurity, suggesting that individual differences in emotional processing may modulate vulnerability to deep fakes and related threats. Future work should further explore these cognitive mechanisms and consider user characteristics when designing AI technologies.
Event Type
Poster
TimeWednesday, October 15th5:30pm - 6:30pm CDT
LocationRiverside East