Case Study 35-2: Facial Recognition at the U.S.-Mexico Border

Biometric Surveillance, Immigration Enforcement, and the Limits of Accountability

Background: Biometric Entry-Exit and Border Surveillance

The U.S. Department of Homeland Security's "Biometric Entry-Exit" system represents one of the largest facial recognition deployments in the United States. Required by Congress (the Intelligence Reform and Terrorism Prevention Act of 2004) but implemented slowly, the system was substantially accelerated after 2017 and is now operational at all major U.S. international airports and many land and sea ports of entry.

The system works as follows: travelers entering or departing the United States have their face photographed and matched against DHS databases — primarily passport and visa photos, but also databases of people with prior immigration violations, criminal records, and watchlists. For U.S. citizens arriving from abroad, the match confirms identity. For foreign nationals, the match also confirms visa validity and immigration status.

The implementation has expanded significantly beyond its original authorization: - Commercial partnerships: DHS has partnered with airlines and airports to use government biometric databases for commercial boarding — travelers can board flights by face scan rather than presenting a boarding pass - Inward expansion: The system has been extended to internal checkpoints, transit hubs, and local law enforcement through information-sharing programs - Database expansion: The databases searched include not only passport and visa records but fusion center data, state criminal records, and other law enforcement databases

Who Is Affected and How

The border biometric system affects different populations differently:

U.S. citizens arriving from abroad: U.S. citizens have the right to enter the country and cannot be denied entry. The facial recognition match is primarily an identity verification step. DHS has offered opt-out for U.S. citizens at some airports (though not prominently advertised), but a 2017 DHS document stated the department's goal of capturing biometrics on 97% of travelers.

Non-citizen entrants: Foreign nationals have significantly reduced constitutional protections at the border. Courts have upheld extensive warrantless searches of foreign nationals at entry points; facial recognition adds to an already extensive biometric processing regime that includes fingerprints and iris scans for most visa holders.

People of color and immigrant communities: The intersection of facial recognition inaccuracy (lower accuracy for darker-skinned individuals) and the immigration enforcement context creates particular risks. A false positive match to a watchlist or a person with immigration violations could trigger enforcement action against an innocent person.

Undocumented immigrants: People attempting to cross outside official ports of entry face a different set of surveillance technologies — ground sensors, aerial surveillance, camera networks, and in some contexts facial recognition deployed by Customs and Border Protection (CBP).

Asylum seekers: People seeking asylum at ports of entry are biometrically processed as part of their application. This data is retained in databases and shared with other law enforcement agencies.

The Constitutional Question: Fourth Amendment at the Border

The Fourth Amendment's application is significantly reduced at the border. Courts have held that border searches — including of electronic devices, which can contain thousands of intimate documents — are generally warrantless and constitutional as "border searches." The rationale: the United States has a compelling interest in controlling what enters the country, and the border is a "functional equivalent" of a border where normal Fourth Amendment expectations don't apply.

The border search exception has been extended to a "Constitution-free zone" that some courts and advocacy organizations have identified: the Fourth Amendment provides reduced protection within approximately 100 miles of any external boundary of the United States. This "border zone" includes approximately two-thirds of the U.S. population, including entire states.

Within this zone, CBP has significantly expanded authority for suspicion-less stops and identification checks — authority that intersects with the biometric identification capabilities of facial recognition.

The ACLU has challenged border biometric processing on multiple grounds: - Fourth Amendment claims (that biometric collection and database matching constitutes a search requiring probable cause) - First Amendment claims (that people are deterred from protected political activities when they know they will be biometrically processed at the border) - Equal protection claims (that enforcement is racially targeted)

These challenges have faced significant legal obstacles, given courts' historical deference to the executive branch on border enforcement.

Documented Abuses and Errors

False positive consequences at the border: While comprehensive statistics on facial recognition errors at U.S. borders are not publicly available (CBP doesn't publish error rate data), documented cases include: - Travelers detained for hours due to database matches to wanted persons who looked similar but were clearly different individuals on closer examination - People with common names having their entries flagged due to database matches to others with the same name and similar faces - Citizens with records in other countries (foreign national databases sometimes merged) being subjected to enhanced screening

Racial profiling: Government watchdog reports and civil liberties investigations have documented that CBP's enforcement activities disproportionately affect people of color, particularly Latino/a travelers along the southern border and Arab and Muslim travelers at international airports. Facial recognition's lower accuracy for darker-skinned individuals intersects with this documented pattern of racially targeted enforcement.

Chilling effects on political activity: Multiple human rights workers, journalists, and activists have reported being subjected to enhanced screening, prolonged questioning, and biometric processing when entering or leaving the United States. The ACLU has documented cases where international travelers were questioned about their political views, organizational affiliations, and journalistic work. Knowledge of biometric processing creates chilling effects on people who work with sensitive populations, speak with government critics, or engage in political activity that governments might wish to monitor.

Clearview AI and Immigration Enforcement

In 2019, Buzzfeed News and other outlets reported that ICE (Immigration and Customs Enforcement) had contracts with Clearview AI. The use of Clearview's web-scraped face database in immigration enforcement raised distinctive concerns:

No consent, no knowledge: People in Clearview's database never consented to have their images used for immigration enforcement. Someone who posted a photo to social media for friends and family could have that photo used to identify their relatives, document their location history, or support an immigration enforcement action.

Community chilling effects: In immigrant communities, fear of surveillance creates significant chilling effects on civic participation, medical care, and reporting of crimes. If community members believe facial recognition is being used by immigration enforcement, they may avoid public spaces, medical appointments, schools, and police contact — to the detriment of community health and safety.

Deportation as consequence: Unlike criminal justice false positives — which can be challenged in courts with constitutional protections — immigration enforcement operates under a different legal framework in which the consequences of a false positive (detention, deportation) may be imposed before adequate review.

Advocacy and Policy Responses

ACLU litigation: The ACLU's projects on border and immigration rights have challenged facial recognition use in immigration enforcement through litigation, public records requests, and legislative advocacy.

Immigrant rights organizations: Groups including the National Immigration Law Center, RAICES, and local immigrant rights organizations have documented surveillance-related chilling effects in immigrant communities and provided advocacy for legislative restrictions.

Congressional action: Some members of Congress have attempted to restrict CBP's use of facial recognition through appropriations riders and standalone legislation, with limited success.

Local action: Some cities have declared themselves "sanctuary cities" with respect to ICE enforcement cooperation. These policies do not directly address federal biometric systems but reduce local cooperation with immigration enforcement.

International law: The UN Special Rapporteur on the human rights of migrants has expressed concern about biometric surveillance of migrants and asylum seekers, arguing that such surveillance must be proportionate, necessary, and non-discriminatory under international human rights law.

Analysis Questions

1. Courts have upheld the "border search exception" to the Fourth Amendment, allowing warrantless searches of devices and biometric collection at the border. Is this doctrine appropriate for the biometric surveillance context? Does the scale and permanence of facial recognition databases change the constitutional analysis?

2. The border biometric system was authorized by Congress for security purposes. It has been expanded through commercial partnerships (airline boarding), inward extension (100-mile zone), and database expansion. This is function creep. What governance mechanisms could prevent such expansion? Why have existing mechanisms failed?

3. Immigrant communities report chilling effects from border surveillance — avoiding public spaces, medical care, schools, and civic participation. How should policymakers weigh these chilling effects against security interests? Is there evidence that border biometric surveillance produces sufficient security benefit to justify the community harms?

4. Clearview AI's use in immigration enforcement relied on images scraped without consent from social media. The people in those images never agreed to be part of an immigration enforcement database. Is this a categorical wrong regardless of immigration enforcement effectiveness? What legal remedy, if any, should be available?

5. The Gender Shades accuracy findings are particularly concerning in the border context: less accurate systems for darker-skinned people, deployed in contexts that disproportionately affect people of color, with significant enforcement consequences for false positives. How should accuracy disparities be factored into decisions about deploying facial recognition in immigration enforcement contexts?

6. The 100-mile "Constitution-free zone" doctrine means that two-thirds of Americans live in areas where CBP has expanded authority. Does this represent an acceptable trade-off between border security and civil liberties, or has the doctrine extended too far? What limiting principles would restore appropriate Fourth Amendment protection?

This case study connects to Chapter 35 Section 35.3 (government uses) and Section 35.6 (wrongful consequences). It connects backward to Chapter 6 (national security) and Chapter 9 (mass interception) and forward to Chapter 36 (racial surveillance) and Chapter 38 (predictive AI).