The United States Customs and Border Protection will launch a “Pedestrian Border Experiment” that will target non-US citizens. The “experiment” is expected to mark the first time facial and iris scans are used on individuals crossing the U.S.-Mexico border.

The San Diego Tribune reported, “Prompting the experiment is an effort by U.S. authorities to track nonimmigrant visa holders who enter the United States legally but remain in the country after the permits have expired.”

It will target “unauthorized immigrants,” people who have “entered the country without valid documents or arrived with valid visas but stayed past their visa expiration date or otherwise violated the terms of their admission.”

The pilot program will occur over a two-month period. It will take place at Otay Mesa, San Diego, where the federal government is likely to scan around 6,000 or more people each day as they cross the border on foot.

Pete Flores, the director of CBP’s San Diego field office, told the Tribune the experiment would process northbound “foreign pedestrians” through kiosks, which would take iris and facial scans. U.S. citizens would not face any changes in how they are processed.

When going to Mexico, non-U.S. citizens will have RFID-enabled documents such as their U.S. visa card.

“As they walk through, we will have an RFID reader, and we will have a device that will read the iris. The intent is to capture individuals while they’re on the move,” Flores said.

According to the Tribune, the area will also have “facial biometric cameras,” and the experiment will “test the viability of facial and iris image capture in an outdoor land exit environment.”

In July, CBP announced it would collect biometric and biographic data from “foreign travelers” departing the U.S. from ten identified airports. This was designed to help identify individuals who had overstayed their visas.

CBP planned to evaluate the test based on “the occurrence of watchlist matches based on biometric data” and the “occurrence of law enforcement hits, including those requiring referral to secondary inspection.” It would last one year.

Both pilot programs were designed to help establish an entry and exit program for non-U.S. citizens that would incorporate biometric scanning.

The security benefit of expanding biometric scanning at the southern border may be very small. The Bipartisan Policy Center concluded in a report [PDF], “Biometric identifiers have greater potential for accuracy than biographic, but this benefit has not been proven in real-world settings.”

“Exit records offer little value for overstays who come into contact with law enforcement—the dominant way unauthorized immigrants are currently identified and removed,” the Center added.

However, it would mean private contractors would find more business opportunities in the ever-expanding U.S. security industrial-complex.

The effectiveness of biometric scanning in any entry and exit program will depend on whether the software works. The Boston Globe reported in 2011 that the Massachusetts Registry of Motor Vehicles had misidentified around a thousand people each year while using facial recognition technology.

Notably, CBP wants to collect data from individuals’ eyes as well as their faces. The Electronic Frontier Foundation has warned against this kind of over-collection. EFF has also pointed to how “inaccuracies” might be perpetuated.

Immigration and Customs Enforcement (ICE) has operated a program called “Secure Communities,” which is a prime example of what can go wrong.

According to EFF, the following process is followed:

ICE’s Secure Communities program shows how a program with a stated purpose to remove “those who pose a danger to national security or public safety” can easily devolve. Under Secure Communities, when state and local law enforcement collect fingerprints from arrestees and run them against the FBI’s [Integrated Automated Fingerprint Identification System] database, IAFIS automatically shares that data with DHS’s US-VISIT program and IDENT database to check the arrestee’s immigration status. If US-VISIT finds either an indication that the person lacks lawful status or finds a “no match,” ICE will issue a detainer on the person until the agency can take him into custody. As such, through data-sharing, ICE is able to conscript state and local law enforcement to enforce immigration laws.

This has been the impact of the program:

According to government statistics, there have been approximately 155,800 deportations since Secure Communities began in 2008. In fiscal year 2011, out of the 79,797 people who were deported under S-Comm, 20,568 were never convicted of Level 3 offenses—crimes punishable by less than one year, including driving without a license. In California alone, there have been more than 70,640 deportations, since Secure Communities started in 2008, and these have included victims and relatives of victims of domestic abuse and people arrested under mistaken identity. ICE does not wait until a person has been convicted of a crime to begin deportation proceedings. According to law enforcement agencies, Secure Communities has had negative effects on community policing efforts.

In other words, there is a strong likelihood refugees will become victims of new scanning protocols. (And, already, the government has lost in court because they are keeping refugee mothers and children in detention centers when they should not be in such facilities.)

EFF has additionally warned that records from biometric scans will be combined with other records used for criminal or national security investigations. There are very few standards, mechanisms for oversight, and little transparency. Data from local fusion centers and private security guards, who file “Suspicious Activity Reports” (SARs) can racially profile migrants or refugees and then those SARs can combine with biometric data, which may lead to deportation.

It is entirely possible that the government will struggle with an outdoor land environment. According to EFF, “Facial recognition’s accuracy is strongly dependent on consistent lighting conditions and angles of view. It may be less accurate with certain ethnicities and with large age discrepancies.”

Finally, any program started to target immigrants will become a template that can be expanded to target U.S. citizens. It could lead to iris and facial scans at polling places, when renting houses, purchasing guns, obtaining credit, boarding planes, attending sports, entering amusement parks, or logging on the internet.

The government can begin by keeping records on all non-U.S. citizens who cross the border. After that process is perfected, government will inevitably determine other classes of individuals worth monitoring need to be incorporated into the process as well.