The last day to post comments on the largest expansion of facial recognition in U.S. history is tomorrow. While 60 million U.S. citizens will be impacted by this new rule, under 50 have reportedly commented on the issue. According to the proposal, the expansion in facial recognition technology use would allow U.S. Customs and Border Protection (CBP) to screen foreign travelers coming in or out of the U.S. using a collection of facial images and other biometrics. The rule would allow for officials to scan travelers not only at airports but also at seaports and land border crossings.
CBP published a notice of the proposed rule in the Federal Register on Nov. 19, 2020. While comments were initially closed, CBP reopened the comment period for an additional 30 days, to close on March 12. Organizations like the American Civil Liberties Union (ACLU) have argued against this new rule because it disproportionally harms immigrants and people of color. The ACLU has argued that facial recognition technology does not work well and even if it did, it impacts the civil liberties and privacy of individuals. The rule proposed during Trump’s administration is expected to be fully enforced by 2024.
“This plan is unjustified, unnecessary, and dangerous. Unlike fingerprints and many other biometrics, faceprints can be collected covertly, at a distance, and without our consent. Once a government acquires a person’s faceprint, it creates a risk of a unique and unprecedented form of persistent surveillance, one that allows the government to identify and track people without their knowledge,” the ACLU said in a statement.
“Congress has not authorized the government to take such an extraordinary and unprecedented step, and the government already collects fingerprints from non-U.S. citizens entering the United States, undercutting its claims about the unique value of facial recognition for identity verification. Face surveillance gives governments, companies, and individuals the power to spy on us wherever we go, and we are already seeing its harmful consequences,” the statement continued.
While facial data collected and algorithmically compared to a database of existing traveler images has already been in use by CBP in a pilot program deployed at select border crossings, this expansion would allow the data collection to expand further and stay in the Department of Homeland Security’s (DHS) system for up to 75 years. In addition to CBP and other DHS agencies, it will be accessible to international governments as well as federal and local authorities, including local law enforcement, according to the notice. As a result, there would be a dramatic increase in the number and use of faceprints collected by the government.
It’s important to note that because this expansion allows other agencies to use facial recognition data, faceprints collected “could enable systematic surveillance by other government agencies and foreign governments.” It would ultimately “expose where people go, who they associate with, and even what they believe, based on the religious services, protests, or meetings they attend,” the ACLU said in a filing.
The ACLU’s filing also detailed that similar programs were used by the government of China, in which police identified members of an ethnic community, Uyghur Muslims, in order to isolate and subject them to human rights violations. In the U.S., government agencies have already been criticized for using similar technologies to track protesters and journalists. Expanding it would make matters of privacy worse.
Not only is facial recognition an issue for privacy, but it disproportionately affects Black and brown people. In several incidents, Black people have been wrongly arrested due to facial recognition technology failing to appropriately identify them. Racial algorithmic bias has been found in multiple fields where artificial intelligence technology has been used. In the past, big corporations like Microsoft, IBM, and Amazon have been called out for using facial recognition technology that was biased against people with darker skin tones. These companies failed to accurately identify people with darker skin and connected members of Congress to those who broke the law.
These and other examples shed light on a wider issue of the lack of inclusion of people of color in developing technology and the bias such databases enforce.
Additionally a study from the Georgia Institute of Technology in 2019 entitled "Predictive Inequity in Object Detection” found that artificial intelligence recognition systems in self-driving cars may have more difficulty detecting pedestrians with dark skin than those with light skin.
Many technological systems use a dataset of images of people that are divided into two groups using the Fitzpatrick scale, a system that classifies human skin tones from light to dark into six categories, Business Insider reported. While image detection models often correctly detected light-skinned groups, they failed to detect those in dark-skinned groups. Bias in facial recognition goes beyond enforcing the law—it impacts the lives of people of color since it puts them more at risk of injury.
CBP should not and does not need to expand the use of facial recognition tech. Not only is facial recognition technology flawed, but it creates the risk of tracking and spying on communities in addition to profiling Black and brown communities and religious minorities. CBP’s history of human rights abuses and detaining individuals is proof that it will not be difficult for the agency to abuse its power with such technology. We must speak up against laws that propose a xenophobic and anti-immigrant agenda. Visit the Federal Register website before March 12 to add in your comments.