[3Xin2] Poster session 1

[3Xin2-41] Artificiality is perceptually associated with trustworthiness but not attractiveness in AI-synthesized faces

〇Jiaxin Wang¹, Satoshi Nishida² (1.Graduate School of Frontier Biosciences, Osaka University, 2.Center for Information and Neural Networks (CiNet), Advanced ICT Research Institute, National Institute of Information and Communications Technology (NICT))

Keywords:Synthetic faces, Artificiality, Trustworthiness, Attractiveness, Perception

How does the perceived artificiality of AI-synthesized information affect human evaluations of that information? Previous studies have demonstrated that the perceived artificiality of synthetic faces is correlated with both attractiveness and trustworthiness. However, although attractiveness and trustworthiness are also known to be correlated, previous studies did not assess the effect of artificiality after controlling for the correlation between attractiveness and trustworthiness. To address this, we collected ratings of perceived artificiality, attractiveness, and trustworthiness for synthetic faces in psychological experiments, and examined correlations between each pair of variables while controlling for the third variable. We found that perceived artificiality showed a clear negative correlation with trustworthiness, while it did not significantly correlate with attractiveness. This result suggests that, at least in facial perception, the perceived artificiality of synthetic information affects social evaluations more than aesthetic evaluations.