By The FirstA ground-breaking study out of Denmark has made fascinating discoveries about how artificial intelligence perceives political ideology and emotional states from facial photos.
Published in the distinguished journal ‘Scientific Reports,’ the study found that artificial intelligence can predict a person’s political leanings with a striking 61% accuracy, simply by analyzing a single headshot.
The researchers utilized about 3,200 publicly available photos of candidates who competed in the 2017 Danish municipal election. The photos were run through Microsoft Azure’s Face API tool, a leading-edge technology capable of assessing an individual’s emotional state from photographic input.
The results unveiled a compelling pattern. “For females (though not males), high attractiveness scores were found among those the model identified as likely to be conservative,” read the findings. This statement complements earlier research using human raters, which also suggested a link between attractiveness and conservatism.
Interestingly, male politicians were not excluded from the phenomenon. The findings indicated that left-leaning male politicians tend to exhibit more neutral, less happy faces than their conservative counterparts. This draws a captivating parallel with the pattern observed among female politicians, presenting a consistent correlation between conservatism and perceived happiness across genders.
“Attractiveness was not the only correlate of model-predicted ideology,” the research team elaborated. “We also found that expressing happiness is associated with conservatism for both genders.”