Last year, in mid-November, U.S. Customs and Border Protection (CBP) proposed a plan to dramatically expand the use of facial recognition technology at all entry/exit ports in the U.S. The plan proposes the use of biometric facial verification to validate the identities of all non-U.S. citizens entering and exiting the country, in which traveler’s “faceprints” are collected and compared to an existing database of images. The plan also specified that traveler images would be stored in a database for up to 75 years and would be made available to international governments as well as U.S. federal and local authorities.

The American Civil Liberties Union (ACLU) delivered a swift and stinging rebuke decrying the new proposal, deeming it the beginnings of a “dystopian surveillance infrastructure that threatens our rights to privacy and anonymity, and disproportionately harms people of color and immigrants.”

The Perils of Facial Recognition

In its proposal, CBP argued that the implementation of an integrated and comprehensive biometric verification system is an optimal solution to protect against terrorism, fraudulent travel documentatio, and overstaying non-citizens. In addition, the plan emphasized that national security was a top priority and that following 9/11 there has been a special impetus for creating a system that screens individuals at multiple checkpoints in their travel journey.

CBP has defended their selection of facial recognition—as opposed to other forms of biometric identification, such as fingerprint scanning—as the best technology as it is “accurate, unobtrusive, and efficient,” and is especially effective for the contactless COVID-19 era. The ACLU, however, warns against the adoption of widespread facial recognition technology since “faceprints can be collected covertly, at a distance and without our consent,” creating a surveillance state in which anyone can be unknowingly tracked the moment they step into public.

Furthermore, the ACLU objects to this drastic expansion of facial recognition technology due to its well-documented biases against people of color, particularly Black people. The seminal 2018 “Gender Shades” study revealed that while the overall accuracy rate of facial recognition was impressively over 90%, that accuracy dipped significantly for Black people, women and people aged 18-30, with there being a 34.4% error rate difference between lighter males and darker females. This spells potential civil rights disaster as misidentifications or failed identifications of certain groups of people have significant impacts when screening for potential national security threats.

It could very well mean that people of color and immigrants would face a lifetime of extra screening and potential harassment globally, particularly considering the data is intended to live in the system for 75 years. China recently demonstrated the dark side of surveillance via facial recognition, with reports of the technology being exploited to detect Uighurs.

The ACLU also pointed out several of CBP’s previous moral blunders, including family separation at the border and dismal detainment center conditions.

Nothing has been set in stone yet, but it’s clear that the debate surrounding this controversial technology is fierce, with powerful entities on either side. The societal and ethical fallout of this decision could affect the world for generations to come.

ABOUT THE WRITER

Lydia You is a computer scientist from Princeton University living in New York City. She is a Tech Innovation Fellow at Identity Review covering the intersection of global tech policy, internet culture and the future of digital media.

Contact Lydia You at lydia@identityreview.com.

Do you have information to share with Identity Review? Email us at press@identityreview.com.