Facial recognition technologies have been found to be more invasive and threatening to privacy than previously thought, according to a recent study published in the journal American Psychologist. Researchers discovered that artificial intelligence can accurately predict a person’s political orientation based on images of expressionless faces. Lead author Michal Kosinski warned that people are unknowingly exposing personal information by simply putting a picture online, and that individuals have less control over their privacy than they may think. This study is a warning about the widespread use of this technology and its potential risks.
For the study, participants were asked to fill out a political orientation questionnaire before an algorithm captured a numerical “fingerprint” of their faces for comparison. The authors collected images of the participants in a highly controlled manner, ensuring that factors such as clothing, jewelry, and cosmetics were eliminated to focus solely on facial features. The facial recognition algorithm VGGFace2 then analyzed the images to determine unique face descriptors, which were used to predict political orientation. The study found that conservatives tended to have larger lower faces, suggesting that facial features can convey a substantial amount of information about a person’s political views.
The authors emphasized the urgency for scholars, the public, and policymakers to recognize and address the risks associated with facial recognition technology to personal privacy. They highlighted that individuals may have less control over their privacy than expected, as stable facial features can convey intimate traits about a person. Kosinski warned that algorithms can be easily applied to millions of individuals quickly and cheaply, posing a threat to personal privacy. The study serves as a cautionary tale about the implications of facial recognition technology that is widely used in various applications.
The study concluded that even crude estimates of people’s character traits can significantly improve the efficiency of online persuasion campaigns. Researchers stressed that policymakers should consider tightening regulations on the recording and processing of facial images to protect individuals’ privacy. The findings indicate that individuals may be unwittingly sharing personal information through their facial features, highlighting the need for increased awareness and safeguards to protect privacy in the digital age. Facial recognition technology poses serious challenges to privacy, and efforts must be made to address these risks in a responsible manner.
