Law enforcement officials across San Diego County, California have run more than 65,500 face recognition scans over the last three years, including thousands of queries by federal agencies such as Immigration and Customs Enforcement, the Drug Enforcement Administration, and the U.S. Marshals. According to records obtained by EFF, the Tactical Identification System (TACIDS) has put 1,309 mobile face-recognition cameras—generally smart phones and tablets—in the hands of investigators, with very little oversight over how the technology is used.
And it’s time for it to stop. Gov. Gavin Newsom recently signed A.B. 1215, a new law that creates a three-year moratorium on biometric surveillance, including face recognition, connected to cameras carried by law enforcement officers. It goes into effect starting Jan. 1, 2020. Now EFF has sent a letter to the body that oversees San Diego’s face recognition program, demanding they shut down TACIDS.
Read EFF's letter to San Diego officials calling for suspension of the face recognition program.
Using technology from the surveillance vendor FaceFirst, TACIDS enables officers to take photos of people in the field—often during routine stops—and matches those images against the San Diego County Sheriff’s database of 1.8-million mugshots. Currently, more than 30 agencies in greater San Diego have access to these devices: officers for various municipalities and universities; county deputies and probation officers; state corrections officers and the California Highway Patrol; and federal investigators, including some surprising ones, such as the U.S. Fish & Wildlife Service and the U.S. Department of Veterans Affairs.
The program is administered by the Automated Regional Justice Information Systems (ARJIS), a joint-powers authority formed by a coalition of San Diego municipalities. ARJIS is in turn governed by the San Diego Association of Governments (SANDAG) and its Public Safety Committee. Although SANDAG’s Public Safety Committee is supposed to review the use policy for this troubling face surveillance program at least annually, but seems not to have not done so since at least 2016, despite significant controversy in the local and national press.
EFF first raised concerns about TACIDS in 2013, after receiving public records about the pilot program. At the time, there were only 133 devices and 67 registered users, who had collectively conducted only 3,058 scans.
Today, there are 1,309 devices and 885 registered users, who conduct more than 25,000 scans per year. In just six years, that’s growth by an order of magnitude in devices, users, and scans. Police can also use the devices to run face recognition on images received via email.
ICE’s involvement in the program raises significant concerns about how local resources might be used for the federal government’s deportation efforts. Between 2016 and 2018, ICE—which currently has 7 devices—ran 309 face recognition scans. That’s about two a week. U.S. Border Patrol has another 2 devices, which ran 53 face scans over the same period. While very little is known about how agencies in the U.S. Department of Homeland Security use the device, anecdotes EFF reported in 2013 show that ICE is enthusiastically using the program. “This is the most awesome thing EVER!!!” one ICE official wrote to SANDAG. Another ICE official discussed how a team detained an individual based on their “spidey sense” that they may be an undocumented immigrant, and then used a TACIDS device to reveal the person’s criminal record.
Unfortunately, under the ARJIS official use policy, each user agency is responsible for investigating misuse by its own members, therefore leaving ICE and the U.S. Border Patrol responsible for their own auditing. In our letter to SANDAG, EFF asks the agency to investigate whether the TACIDS use by these agencies violates either the spirit or the letter of the California Values Act, a 2017 law designed to limit how state resources may be used in federal immigration enforcement.
But it’s not just ICE that raises concerns about use. A presentation on TACIDS delivered at a 2019 conference indicates that San Diego police use face recognition daily when interacting with the homeless population. Although SANDAG says this use is designed to connect people to social services, it demonstrates how emergent police surveillance technologies are frequently deployed among a community’s most vulnerable populations.
A New York Times investigation in 2015 further raised questions about how this San Diego face surveillance technology may impact people of color. Meanwhile, the Sycuan Tribal Police Department used face recognition 770 times. The data shows university police departments frequently use face recognition on campuses.
In passing A.B. 1215, the California Legislature declared that face recognition poses “unique and significant threats to the civil rights and civil liberties of residents and visitors,” and that its use “is the functional equivalent of requiring every person to show a personal photo identification card at all times in violation of recognized constitutional rights.”
With the January 2020 moratorium approaching, SANDAG’s Public Safety Committee must uphold those rights and move quickly to close down this program.