Two recent House hearings have focused on the Department of Homeland Security’s (DHS) use of facial recognition. This comes as Congress attempts to determine how to move forward with facial recognition as its use becomes more widespread.
Recently, it was discovered that Immigration and Customs Enforcement (ICE) is using facial recognition to match driver’s license photo’s with suspected undocumented immigrants. This has come at a time when ICE has made raids around the country, and the Trump administration has faced intense scrutiny for its inhumane treatment of migrants at the border. So far, ICE and the FBI have requested to use facial recognition on state driver’s liscene databases to identify undocumented immigrants in Utah, Vermont, Illinois, and Florida.
Vermont and Utah, had previously allowed ICE to access to the driver’s license database, but stopped allowing ICE access in 2017. The extent of surveillance using state drivers license records according to a joint investigation of Georgetown Law researchers and The Washington Post was more than 390,000 logged cases of searching driver’s license databases in 21 states. Additionally, when doing searches of driver’s license databases law enforcement casts a wide net and collects data of people regardless of their immigration status. This has meant that the wide log of people ICE and the FBI run facial recognition on extends far beyond undocumented migrants.
In Florida the database that ICE and the FBI have access to is known as Face Analysis Comparison & Examination System (FACES) and is broadly used by law enforcement agencies in Florida. The program is part of a greater parternship between local law enforcement and ICE to execute the crackdown on undocumented immigrants.
In the last few years, the use of facial recognition has grown in a number of ways. Cities like Detroit and Chicago have implemented facial recognition on a large, real-time scale. At the same time a number of private companies have begun using facial recognition technology, or developing their own. Delta airlines has begun using facial recognition to verify passengers, a practice that is still optional. The database they use to check on passengers is also linked to Customs and Border Patrol showing how the expansion of this technology in law enforcement has meant greater partnerships with private companies as well. In certain airports, like Charlotte Douglas International, facial recognition could become the norm on international flights, part of an explicit attempt to track migrants coming in and out of the country.
Notoriously Amazon, which collaborates extensively with ICE, developed facial recognition software, “Rekognition,” that proved to be widely unreliable. For example, one study showed the program misidentifying 28 members of Congress as suspects in mugshot photos. In this test, Latino and Black lawmakers were disproportionately misidentified. Amazon had been attempting to pitch this software to law enforcement in an effort to consolidate the market on facial recognition.
The Congressional hearings ended with no clear consensus on the future of facial recognition. Multiple groups have begun petitioning for more stringent requirements on the technology. Defending Rights and Dissent signed a coalition letter of civil rights, privacy and other civil society groups which outlined the issue and encouraged Congress to take action. The letter explained, “The use of face recognition technology by the DHS poses serious risks to privacy and civil liberties, threatens immigrants, broadly impacts American citizens, and has been implemented without proper safeguards in place or explicit Congressional approval. The technology is being deployed today by authoritarian governments as a tool to suppress speech and monitor critics, minorities, and everyday citizens. Congress should not permit the continued use of face recognition in the United States absent safeguards to prevent such abuses.”