For the last 25 years, the U.S. Customs and Border Protection (CBP) has been deploying advanced technology to enhance its ability to keep the country safe and the flow of goods and people moving efficiently through the U.S.’s borders. Facial recognition technology, in particular, is set to become a primary driver of travel and trade in the coming years. According to the World Economic Forum, “up to” 25 percent of international transactions could be facilitated by facial recognition by 2025.
Domestically, the U.S. is not far behind in adopting facial recognition as a primary means of identifying and tracking its citizens. CBP has already scanned more than 21 million people with facial recognition technology in 2020, with the potential for millions more in the future. The agency has also been using facial recognition to identify and track travelers from Canada, northern Mexico and the Caribbean, which have been the sources of the majority of border-related illegal entry apprehensions in the U.S. in recent years. On the Canadian side of the border, the CBP has also used facial recognition to identify and track individuals attempting to illegally enter the U.S. in the hopes of being granted asylum in Canada.
What is Facial Recognition?
Facial recognition, also called biometrics, is the ability of computer systems to identify and track people based on their unique physical characteristics, including their facial appearance, fingerprints, iris, and gait. Facial recognition has become one of the most popular emerging technologies in privacy and security. It’s being used to identify people in photos, track activities, enhance video and audio, detect fraud, and track people’s movements.
How Does Facial Recognition Work?
Facial recognition algorithms compare biometric data — such as a face’s shape, eyes, hair, and mouth — and other biometric data, including gender, age, and ethnicity. This analysis is often done in real time, meaning the computer system can identify people based on photos that were taken only seconds ago.
CBP’s Use of Facial Recognition Technology
CBP has used facial recognition technology extensively since 2012, when the agency began scanning citizens and visitors with the technology at ports of entry (POEs) in Arizona and California. According to an internal CBP report, the agency began using facial recognition at the San Ysidro port of entry in May 2015. The report says the system was installed in response to a surge in illegal immigration along the San Diego border.
According to CBP’s own data, 311 people were initially identified as “match positives” in the first 90 days of CBP’s use of facial recognition technology at the San Ysidro port of entry. The agency also reported that the accuracy rate of the system was 97 percent during that time. The accuracy rate was reduced to 94 percent in December 2015.
Facial Recognition as a Driver of Immigration Control
CBP has used facial recognition technology to identify and track individuals attempting to illegally enter the U.S. in the hopes of being granted asylum in Canada. In February 2016, CBP started scanning the faces of individuals attempting to enter Canada through ports of entry on the northern border. CBP also began using facial recognition at the Detroit/Windsor border in August 2017, in partnership with the Canadian government.
CBP has also used facial recognition to track asylum seekers as they make their way through the immigration court system. In 2017, the CBP started tracking individuals while they waited in custody to appear before an immigration judge. The CBP also monitors the locations of individuals released from custody to track their progress through the court system.
Facial Recognition as a Driver of Border Security
CBP has used facial recognition technology to expand its ability to identify travelers and track their movements at border crossings. Since the agency used facial recognition at the San Ysidro POE in 2015, CBP has scanned more than 21 million people with the technology in Arizona and California. As of mid-2018, CBP was scanning people at more than 30 POEs with the technology. CBP has also used facial recognition technology to track vehicles at more than 300 ports of entry. The agency has also deployed four mobile facial recognition units to the southwest border, which are used to identify vehicles and people at border crossings.
Facial Recognition as a Driver of Trade Control
CBP has been tracking the movement of individuals inside the country with facial recognition technology. In 2018, CBP conducted nationwide, automated scans of travelers at airports and train stations. The agency also scanned people at ports of entry, including at least one biometric entry. All entrants to the U.S. must submit biometrics when entering through an official port of entry.
Facial Recognition of Citizens and Lawful Travelers
Citizens of the U.S. and visitors who are granted permission to enter the country by an immigration officer are not subject to facial recognition scans unless they are also crossing a border. Citizens of Canada and Mexico, for example, are not subject to a biometric entry, including a face scan, when crossing at a port of entry.
Facial Recognition of Unlawful Travelers and Aliens
In February 2016, CBP started scanning the faces of individuals attempting to enter Canada through ports of entry on the northern border. CBP also began using facial recognition at the Detroit/Windsor border in August 2017, in partnership with the Canadian government.
CBP has also used facial recognition technology to track asylum seekers as they make their way through the immigration court system. In 2017, the CBP started tracking individuals while they waited in custody to appear before an immigration judge. The CBP also monitors the locations of individuals released from custody to track their progress through the court system.
Facial Recognition and the Rights of Citizens
Citizens have a right to privacy when it comes to their facial images, and CBP may not create databases of facial images without a valid law enforcement purpose or comply with an order to do so. Given the potential for abuse inherent in facial recognition technology, CBP should continue to proactively disclose its policies, practices, and procedures related to the use of facial recognition technology to ensure transparency and accountability. CBP should also continue to prominently post a notice at all ports of entry, informing citizens that their biometric data — including their face — will be scanned and stored, and how they can request access to their data.
It is important to keep in mind that facial recognition technology is still in its infancy and has obvious limitations. Its accuracy is imperfect and it’s not always able to identify people or distinguish between them. It’s also less likely to produce false positives in a crowd, which could be a concern for a country like the U.S. because of the risk of racial or ethnic profiling. Furthermore, facial recognition technology is not suitable for tracking people across long distances in a consistent manner.
Nonetheless, the potential of this technology is undeniable, and CBP is already using it to expand its control of borders, track travelers, and enforce trade policy.