CBP Eyes Facial Recognition for Vehicles at Border Crossings: Privacy Concerns Mount
The United States Customs and Border Protection (CBP) is once again seeking to bolster its surveillance capabilities, this time focusing on the implementation of facial recognition technology at land border crossings for vehicles. The agency’s latest initiative aims to capture facial images of individuals entering the U.S. by car, comparing them in real-time against existing databases. This move has sparked renewed concerns about privacy and the potential for mass surveillance.
Last week, CBP’s Office of Field Operations Biometric Program Office issued a Request for Information (RFI), seeking vendors to provide technology capable of capturing high-quality facial images of individuals inside vehicles at inbound land ports of entry. These images would be utilized for immediate comparison with the Traveler Verification System (TVS), a pre-existing facial recognition system already in use by the Department of Homeland Security (DHS) at various entry points, including airports and pedestrian border crossings.
Currently, CBP gathers substantial information at border crossings, including a "vehicle package" that encompasses license plate data, scene photos, and biographic information. The proposed facial recognition technology would not replace this existing system but rather supplement it, creating a two-layered approach to vehicle screening.
Under the proposed system, vehicles would first enter what CBP designates as the "Pre-Primary Zone" (PPZ). In this zone, live encounter photos would be compared against images "in government holding." Subsequently, vehicles would proceed to the "Primary Zone," where the new facial recognition technology would capture and analyze photos of any occupants not already "biometrically confirmed" in the PPZ.
According to the RFI, CBP is open to both passive and officer-actuated systems, as long as they provide real-time feedback. The technology must also be capable of filtering out non-human passengers, such as pets or animal graphics on clothing, and capturing usable images of individuals even in challenging circumstances, such as when they are not looking directly at the camera or are wearing hats or sunglasses.
This latest initiative is not CBP’s first foray into the use of facial recognition technology on vehicles. Last year, The Intercept reported on a previous RFI issued by DHS’s Science and Technology Directorate, which sought information on running facial recognition on vehicle occupants as they approached border checkpoints. According to the report, DHS has been testing facial recognition technology on vehicles since 2016, including at the Anzalduas International Bridge, Nogales’ Mariposa Port of Entry, and Buffalo’s Peace Bridge Port of Entry.
In a press release regarding the Buffalo test, CBP stated that it is pursuing facial recognition at vehicular crossings "in accordance with CBP’s Congressional mandate to biometrically record all foreign nationals who enter and exit the United States." The agency also asserted that facial recognition can prevent "imposters" from using other people’s travel documents.
However, concerns remain regarding the effectiveness and accuracy of the technology. The Electronic Frontier Foundation (EFF) obtained a 2022 postmortem report by DHS on the Anzalduas test. While the document stated that all "stated objectives were successfully met," it also revealed that pictures were only obtained approximately 76 percent of the time, and only 81 percent of those images were deemed usable. The reasons for these shortcomings are unclear.
The RFI acknowledges these challenges, stating that "Human behavior, multiple passenger vehicle rows, and environmental obstacles all present challenges unique to the vehicle environment."
CBP’s plans to expand facial recognition at the border have raised significant concerns among privacy advocates. Dave Mass, director of investigations at the EFF, told Wired that CBP’s one-to-one facial recognition system could lead to errors where individuals are not correctly matched to their own documents.
Furthermore, CBP’s ambition to expand real-time facial recognition to vehicles raises a host of privacy concerns. As Mass previously told The Intercept, "We have already seen how automated license plate readers are able to create a massive surveillance dragnet of people’s vehicles and driving patterns."
"If law enforcement is able to add face recognition capture from moving vehicles to the mix, they’ll be able to track not only where your vehicle is going, but who is driving it, and who is in the car with you," Mass continued.
According to the Anzalduas postmortem report, CBP "must significantly increase" the number of images it takes. The RFI appears to be addressing this issue, stating that the new technology would augment "passenger images in the PPZ" and allow CBP to "capture 100% of vehicle passengers." This goal raises further privacy implications, as it suggests an intent to create a comprehensive database of facial images of individuals entering the U.S. by car.
The push for increased surveillance at the border comes amid ongoing debates about the balance between security and privacy. Critics argue that the expansion of facial recognition technology represents a significant erosion of civil liberties and could lead to discriminatory practices. The technology’s accuracy, particularly in challenging conditions, remains a concern, and the potential for misidentification could have serious consequences for individuals.
The RFI underscores the CBP’s commitment to using facial recognition technology as a key component of its border surveillance infrastructure. However, the agency must address the privacy concerns and ensure that the technology is implemented in a way that protects the rights and freedoms of individuals. The potential for misuse and the lack of clear oversight mechanisms raise serious questions about the long-term implications of this expanding surveillance system. The deadline for vendors to respond to the RFI is May 30.
