Ban Facial Recognition Technologies for Children—And for Everyone Else

Abstract

Facial recognition technologies enable a uniquely dangerous and pervasive form of surveillance, and children cannot escape it any more than adults can. Facial recognition technologies have particularly severe implications for privacy, as they can weaponize existing photographic databases in a way that other technologies cannot, and faces are difficult or impossible to change, and often illegal to publicly obscure. Their erosion of practical obscurity in public threatens both privacy and free expression, as it makes it much harder for people to navigate public spaces without being identified, and easier to quickly and efficiently identify many people in a crowd at once. To make matters even worse, facial recognition technologies have been shown to perform less accurately for people of color, women, non-binary and transgender people, children, and the elderly, meaning that they have the potential to enable discrimination in whatever forum they are deployed. As these technologies have developed and become more prevalent, children are being subjected to them in schools, at summer camp, and other child-specific contexts, as well as alongside their parents, through CCTV, private security cameras, landlord-installed apartment security systems, or by law enforcement.

The particular vulnerability of young people relative to adults might make them seem like natural candidates for heightened protections from facial recognition technologies. Young people have less say over where they go and what they do, inaccurate evaluations of their faces could have a particularly strong impact on their lives in contexts like law enforcement uses, and the chilling effects of these technologies on free expression could constrain their emotional and intellectual development. At the same time, some of the harms young people experience are near-universal privacy harms, such as the erosion of practical obscurity, while the discriminatory harms of facial recognition’s inaccurate assessment of their faces are shared by other demographic groups.

The dangers facial recognition technologies pose to human flourishing are insidious enough that a ban on both commercial and government uses is necessary, as more modest proposals will likely be insufficient to counteract its inescapability and discriminatory effects. But children’s heightened vulnerability to privacy violations and discrimination from the use of facial recognition technologies doesn’t diminish the severity of the harms that other groups and the population at large experience. The use of facial recognition technologies on children should be prohibited, and the same goes for their use on everyone else.