“We will not be using facial recognition technology,” South Orange village president Sheena Collum said. “There were a lot of concerns by our residents. The last thing we want is to be using something that disproportionately impacts people of a certain race.”
South Orange began upgrading the village’s street surveillance cameras about two years ago. But, the project was suspended when residents began questioning whether the cameras provided by the vendor would be enabled to allow South Orange police to identify people using facial recognition technology.
South Orange Lt. Adrian Acevedo, the head of special operations for the police department, said there was a lot of misunderstanding surrounding the cameras. The department never intended to install cameras with facial recognition capabilities, he said.
“These cameras don’t tell us who a person is. They don’t recognize a face at all,” he said. “They do not connect to any database that would allow us to identify a person. In plain English, they cannot recognize a face.”
The confusion over the cameras led to a larger community debate about putting policies in place that would ensure South Orange would never enable the devices to use facial recognition technology, local officials said. Collum, the village president, assured residents at a recent village board of trustees meeting that the new cameras being installed in 2023 will not have facial recognition software.
Facial recognition uses biometrics to map the facial characteristics of a person caught on camera. Those features are then entered into a database that can contain millions of other faces in pursuit of a match.
Although the technology is widely used in the private sector for workplace security and to unlock phones, its use by law enforcement is controversial. Critics say facial recognition is an invasion of privacy and the software frequently mismatches people, which could make it a dangerous tool in the hands of law enforcement.
“So many New Jerseyans should have this concern about facial recognition,” said Dillon Reisman, an attorney with the American Civil Liberties Union in Newark. “It can be discriminatory; it can be inaccurate and unreliable. And it can be very invasive in terms of privacy.”
The emergence of facial recognition technology is creating tension at a time when New Jersey has promised to overhaul its police tactics amid rising concerns about brutality in the aftermath of the murder of George Floyd. Advocates say the technology is highly accurate and can lead cops directly to suspects.
Critics say the systems aren’t foolproof. They cite a recent study by the Massachusetts Institute of Technology and Stanford University that examined three facial recognition systems.
The study found that facial recognition systems were extremely accurate in identifying people with light skin, but unreliable with black faces, particularly women, where the error rate ranged between 20% and 34% on the three systems. There’s also been numerous allegations of false arrests attributed to facial recognition, including one involving a Paterson man, Nijeer Parks, who alleged in a lawsuit he spent 10 days in jail in 2019 after Woodbridge Police arrested him for a crime he didn’t commit.
Some states, including Vermont and Maine, have already adopted laws barring law enforcement agencies from using facial recognition. In New Jersey, the ACLU, the state Office of the Public Defender and several Black Lives Matter chapters have called for a similar ban, but the technology remains legal and largely unregulated.
Reisman said it’s not clear how many police departments in New Jersey are using facial recognition cameras, or how many towns have banned them. “Nobody seems to be keeping track,” he said.
The state Attorney General’s office has not adopted a policy on facial recognition systems, so law enforcement agencies are free to use the technology as they see fit.
“There is no general prohibition on the use of facial recognition technology by law enforcement in New Jersey,” said Sharon Lauchaire, an spokeswoman for the state Attorney General’s office.
In the absence of any policy the Attorney General’s has offered police departments a few guidelines on the use of facial recognition, including that it should only be used to generate leads when there are no suspects and should not be the basis for prosecution.
So far, the most notable action the AG’s office has taken to limit facial recognition is to ban one system. In January of 2020, then-Attorney General Gubir Grewal banned police department from using Clearview AI, a system that scraped social media platforms daily to build a database that contained millions of photos. That ban remains in effect, Lauchaire said.
Earlier this year, the Attorney General’s office announced that it would solicit public comments with an eye toward adopting a policy on facial recognition. But since the public comment period ended in March, there has been no policy and no guidance issued.
“The Attorney General’s Office is way behind on this,” said Richard Rivera, the police director in Penns Grove and a proponent of police reform. “This is Big Brother kind of stuff. We don’t have facial recognition in my department, and we don’t plan to use it.”
Sen. Nia H. Gill (D- Essex, Passaic) in February introduced a bill that a bill that would bar government entities from using facial recognition. But that bill, S-1715, has yet to be voted on.
Collum said South Orange officials submitted comments about the legislation, but she is frustrated that she hasn’t heard anything since then.
“Quite frankly, I didn’t realize that these advancements in technology had come this far,” she said. “And the government policies are not keeping up with the changes.”
© 2023 Advance Local Media LLC. Distributed by Tribune Content Agency, LLC.
