Predicting Political Ideology Through Facial Characteristics Using Deep Learning

Researchers in Denmark used deep learning algorithms to predict the political ideology of Danish politicians based on their facial characteristics. The accuracy of predictions was 61%. The study found that right-wing politicians were more likely to have happy facial expressions, while women with attractive faces were more likely to be right-wing. The study raises concerns about the threat to privacy posed by deep learning approaches. However, the study has limitations, including the fact that all politicians whose photographs were included in the study were Danish.