19ºF

Facial recognition technology: Rep. Tlaib raises concerns during tour with Detroit police chief

Chief James Craig continues to push for facial recognition technology

DETROIT – The latest blow to Detroit police Chief James Craig's fight to use facial recognition technology comes from one of his most vocal opponents: Rep. Rashida Tlaib.

The congresswoman said Monday night she feels only black analysts should review cases.

"I think non-African Americans think African Americans all look the same," Tlaib told police.

She made the comments after Craig invited her on a tour of the real-time crime center in response to her raising concerns on social media.

The Detroit Police Department has been using facial recognition for nearly two years, but the ACLU and other civil rights groups argue DPD has been using the technology without specific policies in place.

Nearly two weeks ago, the Detroit Board of Police Commissioners approved use of the software, with restrictions.

This has been a hot-button topic with emotions running high on both sides of the debate.

Craig has invited lawmakers, reporters and photographers to tour the real-time crime center. He invited Tlaib in August after they battled about the issue on Twitter. Tlaib accepted the invitation Monday night.

Critics of the technology call it an invasion of privacy. They said using facial recognition technology puts the civil liberties of Detroit residents in jeopardy.

"It's a tool," Craig said. "It's an investigative tool that can help and that can greatly assist this department."

"We're also -- not surprisingly -- concerned about the potential violations of people's rights as it relates to Fourth, First and Fourteenth Amendment rights," said Rodd Monts, with the ACLU.

The software compares people's photos to images already included in a database to try to identify them.

"Facial recognition is only used on a still photograph and it is only used where there is a violent crime," Monts said.

"It's our analysts who go through with tremendous rigor and identify the most probable suspect, if at all," Craig said. "It's not the technology -- it's the people behind the technology. If we just relied solely on the technology and we went to the top match, we would misidentify probably as much as 95% of the time."

Tlaib told Craig during the tour that she feels the department should only employ black analysts as part of the facial recognition team.

One analyst, Andrew, went through specific training to work alongside investigators.

"There was a shooting at the location," Andrew said. "They said they have video of the incident. So we got a copy of the video and saw where we could get the best screenshot of the suspect. Once we run it through the software, it spits out all these mugshots. There is roughly about 178 mugshots that come up. Then it is up to up to us as analyst being the human component to look through and actually identify the person, if we can."

"He then starts to look at other databases," Craig said. "So he'll scan social media. Magically, he has the name of the subject. He has the subject's social media page. He says, 'OK, the social media photo closely resembles this photograph from the shooting.'"

Craig said he feels critics of the technology don't full understand how it works.

"What's problematic for me is I sit here and I'm trying to offer a reasonable explanation of how we use it," Craig said.

Monts said the ACLU doesn't think there is a reasonable explanation.

"There are people that have come out and been quite vocal to the Board of Police Commissioners about their concerns with regards to this technology, and it seems as though the board hasn't taken these concerns seriously, based on the vote that was taken on the policy," Monts said. "That's problematic."

Moving forward, attention will shift to the Detroit City Council and Michigan lawmakers, where bills are already in the works to try to ban the technology, which has been amended. Right now, it's hard to predict what kind of support the bills will receive, experts said.


About the Authors: