As protesters demand an end to police brutality and the coronavirus pandemic sweeps the nation, police departments around the country are using software that can track and identify people in crowds from surveillance footage — often with little to no public oversight or knowledge.

Dozens of cities around the country are using BriefCam, which sells software that allows police to comb through surveillance footage to monitor protests and enforce social distancing, and almost all of these cities have hosted protests against police brutality in the weeks since George Floyd was killed in police custody, BuzzFeed News has found.

Some of the cities using BriefCam’s technology — such as New Orleans and St. Paul — have been the site of extreme police violence, with officers using rubber bullets, tear gas, and batons on protesters. Authorities in Chicago; Boston; Detroit; Denver; Doral, Florida; Hartford, Connecticut; and Santa Fe County, New Mexico have also used it.

Some cities said they were not using BriefCam in conjunction with the protests or the pandemic. The St. Paul Police Department told BuzzFeed News that it has not used BriefCam “to detect social distancing or face masks” or “for crowd detection/monitoring protests.” The department representative did not respond when asked if there is department policy that prevents it.

BriefCam shows the line between contact tracing, policing, and surveillance can be thin — as cities can spend tens of thousands of dollars for powerful technology, with few restrictions on how police can use it.

BriefCam did not return multiple requests for comment or respond to a detailed list of questions.

Oregon Sen. Ron Wyden raised concerns about the software on Friday, telling BuzzFeed News, “This product appears to offer exactly the kind of indiscriminate surveillance that makes so many Americans concerned about government use of facial recognition technology.”

“Congress should step in and block the use of facial recognition by police in this manner until there are ironclad rules to protect against inaccurate discriminatory algorithms and misuse,” Wyden added. “This company should be held responsible for any and all harm that springs from misidentifications as the company is knowingly marketing a product that cannot provide the certainty necessary for law enforcement functions.”

Founded in 2007 by Hebrew University researchers and now owned by camera company Canon, the Israel-based company sells a system called “Protect & Insights” that lets police and private companies filter hours of closed circuit television and home surveillance and create excerpts of a few relevant moments. Protect & Insights has built-in facial recognition and license plate reader searches, and lets police create “Watch Lists” of faces and license plates. The company also said its tool could filter out “men, women, children, clothing, bags, vehicles, animals, size, color, speed, path, direction, dwell time, and more.”

This month, BriefCam launched a new “Proximity Identification” feature, which it marketed as a way to combat the COVID-19 pandemic. The company claimed it could gauge the distance between individuals, detect who is wearing a mask and who isn’t, and identify crowds and bottlenecks. In a brochure, BriefCam said that these features could be combined with facial recognition to determine the identities of people who may have violated social distancing recommendations.

“Advanced multi-camera search powerfully identifies men, women and children in video with speed and precision,” BriefCam claims on its website. “Accurate face recognition can be leveraged to rapidly pinpoint people of interest using digital images extracted from the video or from external sources.”

It’s unclear if BriefCam can do everything it claims. The company is not a registered vendor of facial recognition software through the National Institute of Standards and Technology, which evaluates facial recognition vendors that are operating worldwide. BriefCam representatives have also never conducted any of the University of Washington’s MegaFace tests, commonly used to assess the accuracy of facial recognition software, according to a list of vendors that downloaded a MegaFace dataset as of February 2020, obtained via public record request.

However, the company has made stunning claims about its ability to parse video footage and extract information about people captured on camera. And police departments have a large appetite for what the company claims to offer: “real-time” crime intelligence, and the ability to “react preventively and proactively” to information” and help in ”reducing the crime rate, spurring economic growth, and enabling situational awareness.”

City governments around the country have paid significant amounts of money to use BriefCam. Denver paid more than $250,000 for BriefCam licenses in 2019. Tulare County, California paid over $119,000 for BriefCam. When Springfield, Massachusetts, paid Motorola $1 million for “hardware, software and radio system upgrades” in 2018, BriefCam was part of the package. Motorola is one of BriefCam’s biggest investors, although BriefCam was acquired by Canon in 2018 for a reported $90 million.

Some of the biggest police departments in the country have used BriefCam. The Metropolitan Police Department of Washington, DC, where law enforcement has unleashed some of the most severe violence on protesters in the country, paid for BriefCam in 2015. An MPD spokesperson told BuzzFeed News the department did not currently use it, but did not say when it had stopped.

A spokesperson for Arlington, Texas, which paid $76,000 for BriefCam in 2019, told BuzzFeed News that its police department only used BriefCam for criminal investigations. “We do NOT use the program for social distancing, face mask, to monitor protests, or any other reasons,” the spokesperson said. But there is a department policy that prevents these applications of the technology, according to the spokesperson.

Police departments outside the US have also used BriefCam, including the Abu Dhabi Monitoring & Control Centre, the Paris Police Prefecture, the Indonesian National Police, the Singapore Police Force, and the Royal Thai Police, according to the company. BriefCam’s technology was also used in Brazil at the 2016 Rio Olympics, and the Boston Police Department said they used it to investigate the 2013 Boston Marathon bombing.

BriefCam isn’t always used or managed by a city or county’s local police department. A representative for the Santa Fe County Sheriff’s Department — a county that paid for BriefCam as a part of a Motorola services package in 2016 — said that the county Regional Emergency Communication Center manages BriefCam, not the sheriff’s department. But the RECC is still a law enforcement agency.

Other cities that have paid for BriefCam include Beverly Hills, California; Overland Park, Kansas; Miami Gardens, Florida, according to the company, as well as Lawrence, Kansas, in 2020; Inglewood, California, and Glendale, Arizona, in 2018; and Coral Gables, Florida in 2016.

“Our primary use of BriefCam has been reviewing large segments of recorded video to locate a suspect or find an event,” an Overland Park Police Department representative told BuzzFeed News. “If we need to review 24 hours of video to locate someone or an offense we can usually get through that in about 20 to 30 minutes with BriefCam. This is a huge time saver for the PD making us much more productive.”

San Diego; Mooresville, North Carolina; Lansing, Michigan, and West Hollywood, California, have all considered BriefCam recently, but it’s unclear if they followed through with contracts.

There are currently no federal guidelines restricting the use of video analytics, license plate reader, and facial recognition software offered by companies like BriefCam. Neema Singh Guliani a senior legislative counsel with the ACLU said that city governments often acquire these technologies without public oversight or debate.

“If something is for a public health purpose, like contact tracing, there’s no need for it to be secret,” Singh Guliani said. “The fact that there hasn’t been that proactive transparency I think is a giant red flag, and I think it raises questions about whether it will contribute to the deterioration in trust of public health efforts more broadly.”