Some are questioning whether police departments can effectively use the controversial techonology to identify criminal without violating citizens’ civil rights.
DETROIT, Mich. (CN) — A Michigan man Tuesday filed a federal lawsuit against the city of Detroit over his wrongful arrest, which he says was caused by faulty facial recognition technology.
Robert Williams, a 43-year-old father from the Detroit suburb of Farmington Hills, Michigan, questions in his complaint whether facial recognition technology is too flawed to ensure that innocent people aren’t mistakenly identified as criminals. Williams is Black and was arrested on Jan. 9, 2020, in front of his family on his front yard, accused of stealing watches from a Shinola store in Detroit.
The suit says a detective relied solely on a facial recognition program to obtain a warrant for Williams’ arrest. Based on the technological match, the suit claims, the detective obtained Willims’ expired driver’s license photo and placed it in a photographic array.
According to the suit, the warrant for Williams was “faulty and misleading” because a detective “hid the fact” that the person who picked Williams out of a photographic lineup — a security contractor who was not at the store when the theft occurred — never saw the shoplifter in person.
The case was dropped by prosecutors less than two weeks after Williams’ arrest when they said the officers had relied on insufficient evidence.
Williams wrote an op-ed for the Washington Post last June in which he described his ordeal.
“The next morning, two officers asked if I’d ever been to a Shinola watch store in Detroit. I said once, many years ago. They showed me a blurry surveillance camera photo of a black man and asked if it was me. I chuckled a bit. ‘No, that is not me.’ He showed me another photo and said, ‘So I guess this isn’t you either?’ I picked up the piece of paper, put it next to my face and said, ‘I hope you guys don’t think that all black men look alike,’ he wrote.
“The cops looked at each other. I heard one say that ‘the computer must have gotten it wrong.’ I asked if I was free to go now, and they said no. I was released from detention later that evening, after nearly 30 hours in holding,” he added.
The lawsuit says people of color are up to 100 times more likely to be misidentified using the technology than white men because the algorithms were primarily trained using Caucasian faces.
Kevin E. Early, a criminologist and associate professor of sociology at the University of Michigan Dearborn said in a telephone interview that the skewed algorithms create an unfair database.
“It hasn’t been perfected in terms of shades, hues, colors,” he said. “It’s much more effective when you are looking at people who are not of color versus people who are of color.”
The Boston City Council voted unanimously in June 2020 to prohibit police from using facial-recognition technology, becoming the largest city in the country to do so after San Francisco, which became the first city on the U.S. to ban the technology in May 2020.
“It’s very divisive right now because of all the things going on nationally,” Early said, citing the number of stories about Black people killed in police-related incidents.
In September 2019, the Detroit Board of Police Commissioners voted to approve the Detroit Police Department’s use of the controversial facial recognition technology to help catch criminals.
“This is a great day,” Detroit Chief of Police James Craig said after the vote.
Craig had submitted a revised directive for the use of the technology that addressed several recommendations made by the commissioners.
The directive said the system would be limited to when officers have “reasonable suspicion” of home invasions and violent crimes involving incidents like shootings, sexual assaults and carjacking. The system would not assess individuals’ immigration statuses and would be restricted from accessing live surveillance streaming video or any security camera device. Outside agencies would be allowed access to the information if needed.
Misuse of the system was supposed to be considered major misconduct that requires notification to the mayor, City Council and Police Board of Commissioners within 24 hours after an incident.
Detroit Mayor Mike Duggan published a letter in which he denied the use of the technology to track innocent citizens but said it was fair game for criminal offenders.
“If your loved one was shot and there is a picture of the shooter, wouldn’t you expect the police to use every tool they can to identify that offender? I fully support the technology’s use for that limited purpose,” he wrote.
Professor Early cited a M.I.T. study where light-skinned men where only misidentified 0.8% of the time while dark-skinned women were more than 34% more likely to be matched in error. Early says financially distressed areas would be affected the most by that discrepency.
“Persons of color who are primarily a large portion of the poor in America don’t have the resources to fight law enforcement,” he said.
But David L. Carter, a professor at the School of Criminal Justice at Michigan State University, says that the technology is improving.
“While it may not provide complete accuracy in some cases, it can provide investigators with good leads,” he told Courthouse News in an email. “With that lead, traditional investigative techniques would be used to confirm an offender’s identity.”
“Facial recognition has a high accuracy rate now,” he added. “Like all technologies, it will improve with more research and refinement.”
Both professors agreed that the growing market for surveillance could benefit from more observation.
“This is a multi-billion dollar industry, with a lack of federal regulation,” Early said.
Carter agreed it needs “strong policy limitations to ensure its use is consistent with civil rights protections,” but said the techonology might ultimately be widely accepted.
“When new technologies are introduced there is often reluctance to adopt it for fear that it will be inaccurate or manipulated,” Carter said. “This happened when digital cameras were introduced and critics said the police would manipulate digital photos to create false evidence.”
“Today, digital cameras are the norm — most of us carry one everyday in our smartphones. The same smartphone that many of us unlock with facial recognition technology.”
Williams is seeking damages for “pain, suffering, humiliation, shame, embarrassment, and emotional distress,” a declaratory judgment requiring officers using facial recognition technology to disclose certain information to the magistrate, an injunction prohibiting the city from using the technology as an investigative technique while it misidentifies people at different rates depending on their race or skin tone, and an injunction prohibiting law enforcement from performing a facial recognition search using any database that has images of Williams.
The complaint was submitted by Michael Steinberg, the director of the Civil Rights Litigation Initiative at the University of Michigan Law School and student attorneys. Williams is also represented by attorneys with the ACLU.