No Longer Just a Face in a Crowd: A Push to Ban Facial Recognition in Public Housing

September 30, 2019

With the advent of facial recognition software, an individual can no longer be just a face in a crowd. The use of facial recognition software has become widespread and increasingly popular in society over the past few years. Facial recognition software utilizes a database of photos, such as mug shots or driver’s license photos, to identify individuals recorded in video footage. The software uses complex geometry to map key facial features. Some key facial features are the distance between a person’s eyes and the distance between their forehead and their chin. The software then creates a “facial signature” that can be compared to a database of known faces.
This year, the Detroit Housing Commission and Detroit Police Department have been working together to erect 26 cameras around two high-rise public housing towers. The cameras will stream directly to the Police, and in the event of a 911 call, facial recognition software will be used to identify those involved. The Housing Commission supports this project because of their interest in protecting the elderly community, who make up a substantial proportion of public housing tenants. The Commission spearheaded the use of security cameras because of the area’s high volume of foot and car traffic and the desire to curb crime. Residents of Detroit say the security cameras are a “badge of shame.” Residents have also claimed that the surveillance is an invasion of privacy and voice that “the only ‘crime’ they committed was being poor.”
Facial recognition technology, even those using highly advanced algorithms, still struggle to accurately identify people of color and women. A study found that the software misclassified women as men 19 percent of the time and mistook darker-skinned women for men 31 percent of the time. Nevertheless, these groups are frequently the subject of facial recognition technology. Its use in Detroit, a city that is approximately 80% African-American, will be troublesome. Failure to resolve the software’s inability to deal with ethnicity will lead to African-Americans and minority ethnic people being falsely identified and unnecessarily questioned by police. Opponents to the use of facial recognition worry that constant surveillance by law enforcement is damaging to society. Facial recognition technology and its use in the justice system will require improved technological accuracy and accountability.
In July, Oakland, California became the third city to initiate a complete ban of its government agencies from using facial recognition technology, after San Francisco and Sommerville, Massachusetts. Massachusetts representatives have called for a ban on all facial recognition technology in federally funded public housing. Worried that this technology will “ensnare innocent people while diminishing privacy rights,” Congress is beginning to look into regulating the use of facial recognition software and surveillance. With the inherent risk of errors and biases in current facial recognition software, one can only hope that Congress begins to regulate these technologies in our communities.
Kaity Emerson
September 30, 2019