Skip to main content

AI Weekly: Detroit’s facial recognition battle is about the ends justifying the means

Image Credit: chombosan / Shutterstock

Join leaders in Boston on March 27 for an exclusive night of networking, insights, and conversation. Request an invite here.


This week, after U.S. Rep. Rashida Tlaib (D-MI) took to Twitter to criticize the Detroit Police Department’s use of facial recognition software, the department invited her to visit its Real Time Crime Center so she could see the technology in action. Rep. Tlaib readily agreed to a visit (how could she not), and we’ll presumably get to enjoy Act II of some choice political theater if and when she actually makes such a visit.

Rep. Tlaib’s remarks followed presidential candidate Sen. Bernie Sanders’ (D-VT) proclamation this week that he would ban all use of facial recognition technology in policing as part of his criminal justice reform plan. Tlaib had already planted a flag for herself opposing some facial recognition with the bill she cosponsored with Representatives Yvette Clarke (D-NY) and Ayanna Pressley (D-MA). Called the “No Biometric Barriers 5 to Housing Act of 2019,” it prohibits “the use of biometric recognition technology in certain federally assisted dwelling units, and for other purposes.”

Her public critique of the Detroit Police Department signals that she’s taking on a new front in the battle. But the brief, three-tweet exchange spoke to larger issues around the use of facial recognition technology within police work. And the department’s response to Rep. Tlaib’s gauntlet toss was telling, demonstrating that its position on facial recognition is that the ends justify the means.

If there was any doubt, after Rep. Tlaib moved on from the Twitter thread, the Detroit Police Department tweeted the following quote: “Abolishing the use of Facial Recognition protects only one group of individuals — VIOLENT CRIMINALS, said Police Chief James Craig.” (Capitalization theirs.)

VB Event

The AI Impact Tour – Atlanta

Continuing our tour, we’re headed to Atlanta for the AI Impact Tour stop on April 10th. This exclusive, invite-only event, in partnership with Microsoft, will feature discussions on how generative AI is transforming the security workforce. Space is limited, so request an invite today.
Request an invite

Then Chief Craig started talking to local TV reporters about how effective his department’s high-tech policing efforts are. Three clips made it onto the department’s Twitter feed. Craig looks to be in full defense mode, standing in the Real Time Crime Center, home of the police department’s Project Greenlight, which incorporates real-time video from all over the city. He says what’s missing from discussions about the technology are the victims of crimes. Then he goes into detail about how responsible the department has been, how it’s taken great care to ensure that Project Greenlight and its facial recognition efforts are constitutional, how it has shown the Center to community groups representing many demographics, how it’s taken into consideration all the well-documented problems with misidentification, and how the police don’t actually perform the facial recognition in real time (it’s applied after video capture only if necessary, he said).

He also insisted that this real-time video is not surveillance, a note the mayor of Detroit, Mike Duggan, sounded in an earlier post. Mayor Duggan said that the facial recognition capability is entirely separate from Project Greenlight, which was developed to help keep an eye on traffic intersections without identifying faces. (A Georgetown study draws a different conclusion.)

You can see from Craig’s perspective that the department doesn’t live in a theoretical world of ethical maybes; they work every day in real and gritty environments, trying to reduce crime in a city with a stained reputation. Creating a fully operational policing tool in the form of the Real Time Crime Center, and adding the capability to use facial recognition software when applicable, must have taken monumental effort and political skill.

It doesn’t take any special technology to read the exasperated weariness on Craig’s face as he talks to reporters. In his mind, he’s done everything the right way. He’s not hiding anything.

But Craig has missed the point. Rep. Tlaib knows plenty about facial recognition technology, and what she knows is enough for her to call for its ban.

Hers is not the only voice decrying facial recognition in policing and elsewhere. There’s an increasing appetite for more regulation around the technology, and it’s already verboten in San Francisco; Somerville, Massachusetts; and Oakland, California. Many academics and researchers are loudly unequivocal about their objections to its existence. There has even been bipartisan support in Congress for legislation around the technology.

The ethical and practical problems involved in facial recognition technology have been exhaustively discussed, debated, and explored, including in this publication. Flaws in the technology, such as people with darker skin being identified with significantly less accuracy than people with lighter skin, are appalling and serve to reinforce existing biases and inequality in policing.

So the hometown beef between Tlaib and the Detroit Police Department is a microcosm of a much larger international debate about facial recognition. While improvements to the technology could theoretically address its accuracy, larger ethical issues, such as police abusing the technology to make arrests, control borders, or target specific groups of people, are not going away soon.

VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.

Want must-read news straight to your inbox?
Sign up for AI Weekly