Police board delays vote on facial recognition technology rules

George Hunter
The Detroit News
Detroit police chief James Craig addresses the media regarding use of facial recognition software.

Detroit — The Detroit Board of Police Commissioners delayed a vote on a policy governing the use of facial recognition software amidst privacy concerns during a raucous meeting Thursday marked by some racial tension.

Detroit police have used the technology for a year, guided by standard operating procedures. The board had been scheduled to vote on a formal policy governing use of the technology, but at the start of Thursday's session, board chairman Willie Bell said action was being postponed.

"It’s still under review, so we can’t go forward," Bell said during the meeting. "We’re still working with the chief and the mayor on the issue."

Although the issue was not voted on by the board, several people voiced their concern at Thursday's meeting, both during the public comments portion and before that, prompting Bell to tell them several times to be quiet. At one point, he threatened to have officers remove people from the meeting.

"This trend toward mass surveillance is a gross violation of the privacy of people who live and work in the city," said Janice Gates, one of the speakers at Thursday's meeting. "This city is becoming more and more gentrified by the day, and it’s not a coincidence this technology is being rolled out in cities where minorities live."

Some who addressed the board, and others in the audience made disparaging remarks about white people. After a woman called Mexican-American Capt. Aric Tosqui, a "white boy," First Assistant Chief Lashinda Stair, who is black, responded: "We're all on the same team. It doesn't matter what color you are."

Several in the audience shouted: "Yes, it does," and Bell again ordered them to be quiet.

During a news conference earlier Thursday, police chief James Craig told reporters during he understands the concerns expressed by some commissioners and community members — but, he added, there is misunderstanding about how the software has been used, and how it will be used if the policy is approved by the board.

If the policy isn't approved, he said the program would be scrapped.

One of the biggest misconceptions about the technology, Craig said, is that police will randomly scan people's faces in public places.

"We're not going to be randomly using this," he said. "The only time we use it is to identify individuals who are involved in violent crimes."

The chief said facial recognition technology is only used after-the-fact, when police have already identified a suspect on video. A still image of that suspect would be fed into the software and compared with police mugshots and pictures in the Michigan Secretary of State database, and on social media.

After a suspect is identified, the proposed policy requires police to use other evidence in conjunction with the facial recognition software.

"That by itself is not good enough," he said. "This is no different from a sketch, which police have used for years. It's a tool we use to get violent suspects off the streets. Is there a chance you may arrest the wrong person? Yes. But it won't be used as the sole evidence against someone.

"People are arrested based on sketches all the time," he said. "But if further investigation reveals that's not the suspect, they're released. That's the same with facial recognition."

Craig said both the existing standard operating procedures and the proposed policy have strict checks and balances to prevent misuse of the technology. He said the only time random scanning of faces would be allowed under the proposed policy would be if there's a "credible" threat of terrorism.

"In order to do that, it would have to go up the chain of command, and I would have to approve it," he said. "I'm talking about a credible threat; if someone just calls in and says they're going to shoot up the Thanksgiving parade, that by itself wouldn't be enough.

"But if there's a credible terror threat, I don't think there's a person in this city who would say they don't want us to find the perpetrator quickly," he said.

The chief also stressed that the facial recognition program is separate from Project Green Light, the program that employs high-definition cameras that are monitored in real time. Craig said the cameras used in the initiative don't have facial recognition software embedded in them.

When The News reported plans for the facial recognition software in October 2017, police officials said it would be used in conjunction with the Green Light program, but Craig stressed Thursday the software is not part of Green Light.

“We’ll use Green Light cameras, so that may be where some of the confusion is,” Craig said. “But we’ll use cameras from any source, not just Green Light. The facial recognition software was never going to be installed in the cameras themselves; that would be cost-prohibitive, and there’d be no need to do it anyway, since we’ll use footage from those cameras, along with other cameras.”

Before Thursday's meeting, Bell had said he was satisfied with the proposed policy and planned to vote to approve it. 

"This has been before us for quite some time, and we've had a chance to review it," Bell said. "It's long overdue. We understand the concerns, and we can always revisit the policy later if there are problems. But I'm satisfied there are checks in the policy (to prevent civil rights violations)."

But Police Commissioner Willie Burton said he wouldn't vote to approve the policy, citing privacy concerns.

"There should have been more community weigh-in," he said. "This sophisticated technology should not be forced on the people. I made a motion at an earlier board meeting to make this a ballot initiative, so the people can weigh in on it.

"There are a lot of concerns I'm hearing from the community," Burton said. "For instance, will this technology be shared with (Immigration and Customs Enforcement) agents, or insurance companies?"

Craig said at his press briefing the data would not be shared or used to target people who are in the country illegally.

The chief added facial recognition has helped police solve recent cases, including the May shootings of five members of the LGBT community, three of whom died.

"We utilized the technology to get a violent suspect off the street," Craig said. "Had we not had that technology, this case would likely be a whodunit."

During Thursday's board meeting, 18 people filled out cards to address the board for two minutes each — far more than usual.

Resident Tawana Petty said she was unhappy the board decided to put off the vote.

"This facial recognition technology is being utilized without transparency, and now the minute we get an opportunity to vocalize our concerns, it’s rushed through. This technology will be the largest experiment on black people in the United States."

Several people said facial recognition software has been found to yield more errors for African Americans.

Only one resident who spoke, Thomas Wilson, said he supported the software. He was jeered by audience members throughout his allocated two minutes.

"My civil rights will be violated when one of these idiots picks up a gun and decides to blow my brains out," Wilson said. "My life, liberty and the pursuit of happiness ends when that fool picks up a gun and decides to shoot me. The police cannot win for losing."

During his press conference earlier Thursday, Craig said not enough attention is paid to citizens targeted by violent criminals.

"How come we never talk about the victims?" Craig asked. "We're not talking about using this to scan burglary suspects. We're talking about violent crimes — and if we leverage this technology and do it in a constitutional way, we win."

ghunter@detroitnews.com

(313) 222-2134

Twitter: @GeorgeHunter_DN