70 Million

Why Detroit Might Be the True Test of Whether More Cameras Make Cities Safer

Episode Notes

Reporter Sonia Paul takes us to Detroit, where 80% of residents are Black, and examines the tools, models and methods changing the nature of policing in the city — from the rise of live-streamed surveillance to facial recognition technology. She investigates their impact on residents, and implications for overpoliced communities of color across the country.


Sign up for our newsletter today.

Episode Transcription

Mitzi Miller:            70 million adults in the United States have a criminal record. In Season Three, we’ll explore how our rapidly changing reality is impacting those in custody, and the policies that keep them there. I’m Mitzi Miller.

Mitzi Miller:             Earlier this year, protests erupted after cell phone footage emerged of a Minneapolis police officer kneeling on George Floyd’s neck, killing him. Soon after, images of police deploying tear gas and rubber bullets on demonstrators calling for an end to police violence flooded the airways. And everyday people pointed their cell phones to capture these harrowing videos. 

                                In the months since, communities across the country — large and small — have continued to pour into the streets to call for change, from reforming to defunding the police.

                                Attention to policing in our country has never been greater. 

                                As the push for divestment in law enforcement and reinvestment in schools and housing gains momentum, how to achieve public safety is being hotly debated. 

                                That’s not surprising, after decades of politicians campaigning on “Law and Order.” 

Bill Clinton:             The first responsibility of government is law and order without it people cannot pursue the American Dream, without it we’re not really free.  

Donald J. Trump:   We must always have law and order. All federal crimes are being investigated, prosecuted and punished to the fullest extent of the law. When the anarchists started ripping down our statues and monuments, right outside, I signed an order immediately. 10 years in prison.

Miller:                      But what’s striking now is a shift in the public’s view of how police operate. As cities examine their own enforcement practices, police departments are investing in new technology, changing how they work, and blurring the line between private and public accountability. 

                                And a foundational question emerges: Who, and what, should we trust to keep us safe?

                                That’s the question at the heart of this episode. We’re examining the tools, models and methods changing the nature of policing — from the rise of live-streamed surveillance to facial recognition technology. Can these systems be trusted in a country that overpolices Black people and disproportionately jails poor people? And what does the future of safety and policing look like?

                                We found some answers in Detroit, where nearly 80% of the population is Black. And where use of video surveillance and facial recognition technology by police is not the future. It is the present. 

                                Reporter Sonia Paul takes a look.

Thompson Curtis:  I am 57 years old and I have a brother who's about eight years older than me. And when he sees a green light, he feels safe. (chuckles) He says, “Oh, oh, they have a green light. Okay, let's let's, let's stop there.”

Sonia Paul:             Myrtle Thompson Curtis has lived in Detroit her whole life. She’s referring to the city’s expansive video surveillance program that streams directly to the police department. 

Curtis:                     He notices the green light. And it's tied directly to... he feels... the police, which are there to either harm him or protect him.

Paul:                        Project Greenlight is the name of the network of surveillance cameras. It’s premised on the idea behind private security and promoted as a “public-private-community partnership.” Instead of businesses handing over footage to police as evidence when a crime occurs, they beam their live  footage to police in real-time — all the time. According to Curtis, this is not the same as actual security. 

Curtis:                     Okay, the placebo effect, you can take two pills. One is a real pill. one is a sugar pill, and because you're being told this is an aspirin, “oh, I feel better.” so when my brother sees the green light, “oh, I feel better.” Even though the placebo effect, it doesn't really mean instantaneously, if something happened, the police would show up. You're not really safer, you just feel safer…

Paul:                        As some Detroiters dismiss this placebo effect, police have insisted these cameras can deter crime


James Craig : And I will tell you, I'm excited, because I think we're poised to do something that's not being done anywhere. So I was passionate and continue to be passionate about the Real Time Crime Center.

Paul:                       This is Detroit’s chief of Police James Craig speaking in January of 2016, at the launch of Project Greenlight. According to the Detroit Police Department, participating businesses in this real time crime center invest in cloud storage to preserve video footage they capture for 30 days. Participants also buy flashing green lights that illuminate the area surrounding their property. With this buy-in, these locations then get more attentive police patrols, and, if a crime occurs, response priority.

Duggan:                 So it's a powerful tool….But what makes it even more exciting is a fact that we're now connecting with our business partners…..They made the investment to purchase high quality cameras, lighting, specific bannering.

Paul:                        Before they launched Project Greenlight, the Detroit Police Department toured multiple centralized video surveillance centers for ideas. These places are often referred to as “Real Time Crime Centers.” 

                                The program has grown enormously.

News Clip 1:            Project Greenlight began with gas stations but has now spread to include restaurants, markets, liquor stores and fast food joints. 

News Clip 2:           And with that, the city’s 100th Greenlight business is up and running...

News Clip 3:           Now the Greenlight system has come a long way since 2016, we’re talking about eight gas stations, now up to 500 partnerships...

Paul:                        At the time of this reporting, about 700 Project Greenlight locations existed. Hospitals, restaurants, apartment complexes, nursing homes, grocery stores, churches. It amounts to thousands of cameras across the city — and that’s on top of private security cameras. Some streets have green lights dotting them for miles, while certain businesses have banded together to become “corridors” of green lights


Captain Aric Tosqui: Everyone take down any seejus information, please! 

Paul:                        That’s “seejus” as in C-J-I-S, short for “criminal justice information systems.”

Paul:                        This is Captain Aric Tosqui. He’s a commanding officer with Detroit’s Real Time Crime Center and crime intelligence unit. That’s where Project Greenlight footage live streams at the Detroit Police Department.

Tosqui:                    Whenever we give a tour, we make sure we take any law enforcement sensitive information down.

Paul:                        And, he’s giving me a virtual tour of this Real Time Crime Center on Zoom. Given the pandemic, only a few analysts are there the day I tour because most of them are working remotely. 

                                Two huge screens composed of smaller square and rectangular images dominate the big, dim room. They span nearly the entire length of a wall. 

Tosqui:                    That wall that I'm pointing to over there is for our traffic management center. 

Paul:                        The setup he’s pointing me to is expanding. The chief of police says it’s because of growing demand for Project Greenlight. Under this expansion, two new “mini centers” will also launch. 

Tosqui:                    The whole thing will be one video wall. And we'll have more analysts and more seating for analysts. It’s going to go over that way..

Paul:                        The Electronic Frontier Foundation estimates about 58 Real Time Crime Centers exist in the U.S. But Detroit’s is unique in how involved it is with private business. 

Tosqui:                    If it's a problematic Greenlight location or a place that requires a little bit more special attention, maybe a place that you know, the owners have called and said, “Hey, we were experiencing break ins and you know, could we get on special attention or something like that?” Well, we may put that on the main screen so that way, we got a little bit more eyes on it, (people as they walk by)....


Paul:                       Here’s how it works: Crime analysts sit at stations set up with multiple computer monitors. Each analyst is responsible for a geographic area, and within these areas are numerous Greenlight locations. Each location typically has four to six cameras. The analyst doesn’t monitor all these video feeds at once — that’s not humanly possible — but cycles through them.

Tosqui:                    When they have that time on their shift, they will pull cameras up at a location, watch those cameras for a few minutes and then go to the next location, watch those cameras for a few minutes.

Paul:                       Greenlight partners still have to call 911 in an emergency. And officers patrol both Greenlight and non-Greenlight locations..

Tosqui:                    The distinction is that when they go to a Green Light location, the analysts here are notified of it. And so that way the analysts can provide an extra set of eyes. While the officer’s inside the location, the analysts can watch the outside.

Paul:                        Detroit police spent nearly $8 million to build out its real time crime center, which included buying the video and data monitoring system from Motorola Solutions. The expansion of the center cost another $4 million.

Tosqui:                    We're able to start investigating and doing the intelligence workouts contemporaneous with whatever the detectives or officers are doing on the street...for example, crime occurs at Project Greenlight. We're able to rewind the tape, so to speak and we're able to see, okay, let's say the suspect goes into a location and the suspect is masked up because of COVID. Maybe that suspect wasn't wearing a mask when he or she was walking into that location, we're able to go back in time a little bit, take a look at that, and then relay that information in real time to the officers who are approaching or… the detectives who are investigating. 

Paul:                        It’s also not uncommon for businesses outside of Greenlight to sign memorandums of understanding with law enforcement, and offer their own private video footage to police. But only Project Greenlight locations stream to this Real Time Crime Center, and only Detroit Police are supposed to have ready access to this footage.

Tosqui:                    Anybody else that would need to get something would have to have a warrant.

Paul:                        That includes federal and state agencies, unless they’re part of a mutual task force with the Detroit police and the situation is relevant — say, for example, investigating a homicide at a Greenlight location.


Paul:                        Detroit Police have cited multiple statistics on lowered crime rates in the city as proof Project Greenlight is working. But critics and privacy advocates are quick to note that crime rates have been lowering overall across the country. And a long-term study comparing Greenlight locations with non Greenlight locations has yet to emerge.

                                But someone did do a study on the first wave of Project Greenlight. He says a goal of the program was to increase the connection between business owners and the police.

Giovanni Circo:      So historically the people have been unwilling to call the police or contact the police, either out of...distrust or feeling that if they called up the police...wouldn't show up in a timely manner. 

Paul:                        Giovanni Circo is an assistant professor at the University of New Haven who studies crime, and has researched the Detroit Police Department. 

Circo:                      That's a lot to do with understaffing at DPD...and the fact that the area of Detroit is just really so large.

Paul:                        Circo’s PhD dissertation at Michigan State University compared the first 86 Greenlight partners with businesses that didn’t have Project Greenlight.

                                He found that Greenlight partners ended up contacting the police more.

Paul:                        Which is what Detroit police also say from their anecdotal experience. At the forefront of these calls are crimes businesses didn’t previously often bother to call the police for.

Circo:                      I’m thinking about minor property crimes, people shoplifting from the stores, stealing gas or something like that.

Paul:                       Circo says over time, some of these minor crimes did decrease. 

Circo:                      Obviously, the one thing that people are always interested  about is... did it have an impact on violent crime? 

Paul:                        Which include homicide, aggravated assault, sexual assault, robbery, and home invasion. And what Circo’s study found matches the results of other studies on closed circuit TV cameras.

Circo:                      ...in that it didn't have really a very large impact on violent crime. 


Circo:                      I think one of the...biggest concerns and criticisms that I’ve heard... saying that, Green Light is a kind of like a pay to play system. Where..you know if you want to have..your calls prioritized that you have to sign up for it. And, you know, I think historically, we haven't seen really anything quite like this….? It's interesting,...as a policing researcher, just because a lot of this is sort of uncharted territory, and we're really still in the kind of the opening stages of...what the police are going to be? What are the police gonna look like in 10 or 15 years?.….

Eric Williams:          Um, all right. Have you ever been to Detroit before?

Sonia:                     Unfortunately, no, I would have gone had COVID not hit.

Williams:                 Okay, so hold on. I'm gonna see how this works. Let me try something here…

Paul:                        This is Eric Williams. He’s an attorney with the Detroit Justice Center, a nonprofit law firm in the city. And, he’s taking me on a tour of Detroit. On FaceTime.

Williams:                 So, I'm gonna ask you to take a walk with me here. Believe me. I am aware of the sort of almost self referential demands that COVID has put on folks.

Paul:                        Williams is highly critical of Project Greenlight, and says Detroit Police sold it as a deterrent to violent crime. That’s why he believes some people don’t resist the program.

Williams:                  ...they've been willing...to give up... their privacy and to some degree, their liberty, for what they perceive as safety….

Paul:                        And that “safety” is also rooted in the notion — that’s not proven — that the mere presence of cameras and lights deter crime. Studies of surveillance programs in Washington, D.C., Chicago, and Baltimore have been inconclusive. Research on the long-term effects of Project Greenlight, including whether it’s increased or decreased arrests, had not been released at the time of this reporting. 

                                As we walk around downtown, Williams points out the Greenlight business corridors, and the green decals on the storefronts hosting Greenlight cameras. If it weren’t for these indicators, the cameras would seem no different from private security cameras.

Williams:                 There's a camera...and there are a number of cameras at each location. There's a camera over there.

Paul:                        Given how integrated Project Greenlight is in Detroit, like, how out of your way would it be to avoid Project Greenlight? Or could you do it pretty successfully?

Williams:                 Okay, more cameras up there. No, you couldn't.....People just kind of ignore it, to be honest.

Paul:                        But Williams has no doubt what complacency with technology contributes to.

Williams:                 In the one sense, it's just an extension of the police, which means whatever problems you had with the police, you're now just doing virtually….

Paul:                        Williams says a big problem with Detroit police is their lack of transparency. Especially when it comes to another tool that’s changing how they do their job: facial recognition technology.

Williams:                 It was developed without much input from, you know, primarily poor people and people of color don't play a role in it, um... And, you know, didn't design it. So that is problematic from the jump. And this is all being enabled by the artificial intelligence that is behind all these systems….


Paul:                        The Detroit Police Department started soliciting bids for facial recognition software in September of 2016. It purchased the technology from the South Carolina company DataWorks Plus in July of 2017. It cost a million dollars. At the time, the police revealed it had already been borrowing software from other agencies to investigate violent crimes for more than a year. And once Detroit police had its own facial recognition setup, they started using it without any external oversight.

Williams:                 So, in the absence of that kind of transparency, this, this technology is only going to be only going to be abused. 

Paul:                       The Detroit Police Department has requested over $200,000 from the city to extend it’s facial recognition contract with DataWorks Plus. The Detroit City Council had yet to vote on that request at the time of this reporting. Public pushback on facial recognition in the city skyrocketed this year. 

                                A civilian body that oversees the police, the Detroit Board of Police Commissioners, was only asked to weigh in on the use of facial recognition software for the first time last year — when the police were already a year and a half into its three-year contract with DataWorks Plus.  As the policy came to a vote last year, a loud backlash erupted at a Board meeting.

Tawana Petty:        (Board of Police Commissioners Meeting) Hi, my name is Tawana Petty. 

Paul:                        Tawana Petty has long been researching surveillance and policing.

Petty:                       I'm here representing the Detroit Digital Justice Coalition and Detroit Community Technology Project. Serve as a director of data justice programming, so I'm very well researched on this technology. 

                                My mother is a retired police sergeant, and my sister is a captain. So I'm not here as an individual who is in an “us against them” type of mentality. But I will say this, the system is flawed. The system is racially biased....Inherently, we're in the largest Black Mecca in the United States. And most ahead, you could actually say that half of the folks that are in jail are only in jail because they can't afford bail...And so if you add a technology, a technology that’s been proven all across the United States…

Paul:                       Both government and academic studies have shown that facial recognition technology misidentifies Black people more often than white people. Cities including San Francisco, Oakland, Boston, and Somerville have banned the use of facial recognition by government agencies, including police. 

Petty:                       I'll close with this,


Petty:                       I promise you I'll close with this. 

                                  (Thank you)

Petty:                        If you please.

                                 (Ma’am, your time is up please).

Paul:                       As Petty emphasized the wide-ranging effects of the technology, the room got even more heated.

Petty:                        We got here for two hours. This is my two minutes.

                                 (Please respect the process, ma’am)

                                 (The person for the community. I love to hear her! Speak up, ma’am.)

Petty:                       Civilians, civilians are not going to be the only ones impacted by this technology. So will police officers.

Paul:                        She’s referring to how a Black police officer in Detroit admitted facial recognition technology couldn’t identify him.

Petty:                      This racially biased system sees that even Black officers are profiled, and if you add facial recognition technology on top of that, there is no single person in this society that is safe from this. And I'm pleading with you on my life. 

                               (Thank you. Let the next speaker come forward)


Paul:                        It was after this meeting that Detroit city officials finally confirmed their policy: police — and only the trained analysts among them — can only apply facial recognition software on still images —  whether they’re from Greenlight locations, non-Greenlight locations, or elsewhere like camera phones. Not live video. And only for cases of violent crimes. 

Tosqui:                    Our software will then search mugshots and mug shots only.

Paul:                        That’s Captain Tosqui again. He says if nothing comes up in the search and the case is serious, Detroit police can escalate the image to Michigan state police. And the state facial recognition software can search a database containing driver’s license photos.

                                If the Detroit police department’s software does spit out a number of possible matches based on its algorithm, it’s up to the analyst to dig further to determine who might be a suspect.

Tosqui:                    They'll be able to pull up previous mugshots for that person. Then they'll start looking at open source information from social media. They'll start looking at driver's licenses at that point.

Paul:                        And if the officer finds a match, they have to get another trained analyst to corroborate or refute whether it’s accurate.

Tosqui:                    If that second train analyst corroborates it, then has to go to a supervisor. And then once a supervisor signs off on it, only then can it go back to the detective as a lead only.

Paul:                        In other words, facial recognition is only supposed to be used as one tool in the investigative process.

                                But that explanation contradicts allegations made this year..

Tosqui:                    In the most recent cases that have populated on the news that I'm sure you've read, the detectives haven't done the work that needed to get done.

Paul:                        What he’s referring to are two cases that emerged in the summer that reveal how easy it is to select the wrong face when relying on an algorithm. 

Robert Williams:     When we get to the interview room, the first thing they had me do was read my rights to myself. And then sign off that I read and understand my rights. 

Paul:                        This is Robert Williams speaking in a video the ACLU released. In January, he was arrested outside his home 25 miles from Detroit. 

Williams:                A detective turns over a picture of a guy inside Shinola I know and he's like, so that's not you. I look. I said, No, that's not me. He turns another paper or we say, I guess that's not you, either. 

Paul:                        The images were taken from surveillance videos at a luxury store in midtown Detroit called Shinola. Five watches were stolen from the store in October of 2018. 

Williams:                 I picked that paper up and hold it next to my face. I said, this is not me. Like I hope you all don't think all Black people look alike. And then he says, The computer says it’s you. 

Paul:                        But even the officers could see that they were wrong with the evidence in front of them.

Williams:                ]He laid the papers out on the table. At that time, I was still guilty in their eyes until the pictures don't match. And they left them on the table and they looked at each other like oops. 

Paul:                        Williams was detained in jail for 30 hours and released on bail. His case is the first known wrongful arrest involving facial recognition software. His story captured media attention as the entire country became enraptured with how policing functions in our communities.

                                Protesters in Detroit spoke out against undue surveillance of Black bodies, and demanded an end to the systems they saw as perpetuating that.

Clip:                         Getting rid of…

                                 Getting rid of…

                                 Project Greenlight…

                                 Project Greenlight…

Paul:                        Williams’ case struck a chord. Members of Congress subsequently introduced a bill to stop federal use of and funding for facial recognition surveillance technologies. Black Democrats in the Michigan House of Representatives called for a ban on facial recognition software, citing how it misidentifies people of color. And questions and worries over police technology loomed at another Police Commissioners meeting. Both commissioners and residents voiced their concerns on the public Zoom call. 


Caller 1:                    This is unethical. It is unreasonable. It does not work to prevent crime or make people in this city feel safer and it needs to stop right now….stop the use of pay to  police surveillance programs like Project Greenlight…

Paul:                        Another caller brought up how the city’s facial recognition contract with DataWorks Plus was up for renewal — costing taxpayers money.

Caller: :                   The thing that I'm not clear on is why if the Detroit police department is only using facial rec on still images, why do they need such a robust contract with DataWorks that allows for concurrent surveillance of 100 video feeds simultaneously, which we can only imagine would come through Project Greenlight. Why does it need mobile facial recognition technology that can be used on handheld units by members of the police force? Why do we need all of these extra capabilities…? 

Paul:                        And this commissioner asked the police a question that’s been on many peoples’ minds...

Commissioner:        Do you know if there's any other cases in the hopper of cases that clear violation of the facial recognition, is anybody else that may have been arrested or detained even just for a minute? If so, that would be very troubling also.

Paul:                        Shortly after this meeting, news emerged that revealed another Black man in Detroit had faced a similar situation. Here’s a newscaster at WXYZ in Detroit.

News Clip:              Michael Oliver is at least the second person DPD wrongly arrested due to a facial recognition misidentification, and now that mistake could end up costing taxpayers.

                             Oliver:    I didn’t commit this crime. So why am I going through it?

Paul:                       Oliver was charged with grabbing a teacher’s cell phone, and throwing it to the ground while the teacher filmed a school fight. The teacher had identified Oliver as a match from a six-person photo lineup pooled from facial recognition software. And the police signed off on it — even though Oliver’s photo clearly showed he has tattoos up and down his arms, and the person’s arms in the probe photo had no tattoos.

                                Once the case went to court, the detective in charge admitted he didn’t follow up with any other potential witnesses or leads

Burton Harris:         I don't trust anything that we're being told about this technology, quite frankly anymore. I'm trusting what I'm seeing with my own eyes and experiencing through my clients. And from where I'm sitting, this technology is being used as the sole tool for investigations. 

Paul:                        Victoria Burton Harris is the attorney for Robert Williams — the man whose wrongful arrest prompted renewed attention on facial recognition. 

Harris:                     So this was the first known case and it was not known because they shared it with me. it was known because they….said it in front of my client. That's the only way that I know that the facial recognition technology was used.

Paul:                       Detroit Police don’t actively track facial recognition usage to arrests.... though experts have suggested, based on analysis of reports from the police, that Black residents are disproportionately likely to be searched using facial recognition.

                                And Burton Harris says she and her colleagues routinely ask for evidence of whether facial recognition or Project Greenlight have informed their clients’ arrests. But no case had ever emerged that they knew of that demonstrated their use — until Williams’ case — which Burton Harris became involved with after his wife got the ACLU’s attention. She says the police didn’t even ask her client where he was the date and time of the robbery.

Harris:                     Once they had a match and a hit, using this facial recognition technology, it informed everything else they did subsequently. In their mind they had the right guy and they didn't need to ask the basic of questions.

Paul:                        She says the case could have turned out far differently.

Harris:                     Well, if he were charged with a more dangerous crime there is a strong possibility that he would have been still prosecuted. He also could have faced a jury trial if he chose not to take a guilty plea. And a jury could have used that faulty facial recognition technology and believe that technology is right and science doesn't get it wrong. And they could have convicted for a crime he didn't commit.

Clare Garvie:          We know that there are wrongful convictions. 

Paul:                       This is Clare Garvie, senior associate with Georgetown Law’s Center on Privacy and Technology. She specializes in how U.S. law enforcement use facial recognition software, and its risks to privacy and civil liberties. 

Garvie:                    We know that a number of people take plea deals and plead guilty to crimes they didn't commit because it's easier, or life circumstances have it so that they can't go and fight their case in court. 

                                So the problem here is that we're using a technology that we have no idea how well it works, that leads to real life consequences. And we actually have no way of figuring out how often it's leading to the wrong person being charged, convicted, or taking guilty pleas on the basis of the flawed technology.

Paul:                        The experience of some Detroit lawyers  — of requesting evidence on the use of facial recognition software and receiving nothing in return —  is not uncommon when held against Garvie’s research. Georgetown Law’s Center on Privacy and Technology estimates about a quarter of the country’s 18,000 law enforcement agencies have access to a facial recognition system.

Garvie:                    What we see across the country is that face recognition is not provided to the defense.

Paul:                       Garvie warns that in the national conversation around defunding and divesting from police, we must continue to scrutinize the implications of technology.

Garvie:                   There will be a push, certainly, on the part of companies to market surveillance technologies as a cost effective way for law enforcement to continue doing their job to continue operating at the level they're at, if not a higher level, with lower budgets. But that's not the idea behind divesting in over militarized, racially biased policing. Increased surveillance is going to be disproportionately on communities of color. That is a very quick way to perpetuate existing inequities in the system and to amp up the militarization of police, even while spending less money.

Paul:                        In June of 2020, after Williams’ case drew wide scrutiny, Detroit’s Mayor Mike Duggan at a press conference defended the technology. 

Duggan:                 I think there are serious issues that we're looking at. But to say that reflects on the, the technology, I just don't think it's accurate. It's like saying that because a detective and a prosecutor made a mistake on the handling of DNA in a case you should no longer use DNA. 

Paul:                        But as the entire nation reckons with investments in policing, some emphasize how new tools and systems police use don’t change existing issues in the criminal justice system.

Petty:                      Now we've gotten to the point where these systems are integrating with each other and rooting their determinations based on already biased data that comes from other systems. 

Paul:                        That’s Tawana Petty. The data and technology researcher who spoke out at the police commissioners meeting. She has a different theory about what contributes to safety. 

Petty:                      But if you drive into a suburb where the communities are more wealthy, there seems to be this abundance of imagination... 

Paul:                        And she says there’s an easy way to see this at work.

Petty:                      Folks understand how to resource neighborhoods; how to minimize crime without having a police officer on every corner, and surveillance cameras all up and down light poles all over the neighborhoods, and flashing green lights that look like Scarlet letters that are marking these territories as dangerous, right?

                                Most people have seen Detroit as a hopeless, helpless, violent Black city….if you grow up in a city where you're taught that you have to get out of that city in order to make something out of yourself, and if every time you turn on the news, there's a particular images that are fed into your psyche, if you go into the school systems and you see that they're being disinvested...if you come home, and you see dilapidated buildings...When you see that kind of stuff happening, then it has a psychological impact...

Paul:                        Which is why she thinks some Detroiters are so quick to equate safety with surveillance. But she says her tipping point in refusing to be complacent with that understanding was realizing how rapidly Project Greenlight had expanded.

Petty:                       And it was such a rapid expansion with with very little analysis on whether or not it was even creating safety and very little analysis on what was happening with the data, how long the data was going to be held, what impact the data had at that point had had on community members… and then the...understanding that facial recognition was now a component of that system really heightened the visibility around how drastic the impact could be on communities of color, particularly Black communities…


Rev. Horace Sheffield: 

And so, sometimes people at the bus stop were being robbed right outside of my window, and you can hear it.

Paul:                       Reverend Horace Sheffield the Third is a pastor in Detroit, born and raised in the city. He’s also the executive director of the Detroit Association of Black Organizations. 

Sheffield:                Being in my building every single day and hearing people scream, witnessing crimes being committed at this gas station across the street, it had become really the bane of our community.

Paul:                       Back in April of 2017, the Detroit Association of Black Organizations convinced a gas station owner to install Project Greenlight. 

Sheffield:                We paid for half of it to happen…. I mean people really felt as if they can come back out and reclaim this space that had been stolen from them.

Paul:                        He says when it comes to deciding whether we should lean on this technology, we have to make sacrifices.

Sheffield:                I'm not telling you, I'm 100% you know, comfortable with all of it. But I certainly know based on my experience that has had a tremendous impact, and therefore I'm willing to take some risk based on the rewards….I do have a problem with cameras everywhere, capturing us all the time, doing no matter what.

Paul:                        But at the rate Project Greenlight is growing, there will be cameras everywhere. There already are thousands of cameras in Detroit, managed by both police and private security. And facial recognition remains in the city, for now.

                                Some communities are promoting an initiative called “Green chairs, not green lights,” as a way to reframe what surveillance should look: People sitting out on their front porches and looking out for each other… 

Myrtle Thompson Curtis is helping to organize the initiative.

Myrtle Thompson Curtis: 

Our goal is to strengthen the use of the chair and to strengthen our community and change the culture of policing and safety through this action.

Paul:                       Cities like Detroit are only beginning to understand the full scope of how these technologies affect communities, how local police use them, and how they interact with state and federal agencies. If they continue to adopt them, research and analysis of their impact, as well as local and national legislation, will need to keep pace. 

                               For 70 Million, I’m Sonia Paul.

Miller:                     Since recording this episode, Michael Oliver, who’s arrest stems from being misidentified by facial recognition technology, is now suing the teacher who picked his photo from the lineup, the police officer who issued a warrant for his arrest, and the city of Detroit for injuries and damages. According to the lawsuit Oliver suffered degradation, humiliation, emotional distress, pain and suffering due to physical injury, and economic loss among other damages.

                                Thank you for listening. For more information, toolkits, and to download the annotated transcript for this episode, visit 70millionpod.com. 70 Million is an open-source podcast, because we believe we are all part of the solution. We encourage you to use our episodes and supporting materials in your classrooms, organizations, and anywhere they can make an impact. You may rebroadcast parts of or entire episodes of our three seasons without permission. Just please drop us a line so we can keep track. 

                                70 Million is made possible by a grant from the Safety and Justice Challenge at the MacArthur Foundation.

                                70 Million is a production of Lantigua Williams & Co. Season 3 was edited by Phyllis Fletcher and Laura Flynn, Cedric Wilson is our  lead producer and sound designer. Virginia Lora is our managing producer. Leslie Datsis is our marketing lead. Laura Tillman is our staff writer, and Michelle Baker is our photo editor. Sarah McClure is our lead fact-checker. Ryan Katz also contributed fact checking. Juleyka Lantigua-Williams is the creator and executive producer. I'm Mitzi Miller.


Paul, Sonia. “Why Detroit Might Be the True Test of Whether More Cameras Make Cities Safer.” 

70 Million Podcast

, Lantigua Williams & Co., September 28,, 2020. 70millionpod.com