‘It’s humiliating’ – When a faulty facial recognition system led to arrest of innocent Black man

In what is said to be the first known case of its type, a Black man in Detroit was arrested in front of his wife and two daughters after being mistakenly identified as a thief by a facial recognition algorithm. Robert Julian-Borchak Williams, of Michigan, was arrested earlier this year on charges of theft as the police claimed his face was matched with CCTV images from a robbery.

It later turned out that the software had mistakenly identified two Black men as the same person. 

The NPR reports that 42-year-old Williams, after his arrest, was sent to the station where detectives showed him a blurry still from a surveillance video of a man shoplifting from a boutique store in downtown Detroit.

“Is this you?” the detective asked him. “No, this is not me,” Williams replied. “You think all black men look alike?”

The security camera image showed the suspect wearing a red St Louis Cardinals hat. Williams, who works in an automotive supply company, said he would in no way be wearing that kind of hat.

The detective subsequently held the image up to Williams’ face and agreed: “I guess the computer got it wrong,” he said, according to NPR. Realizing they had made a mistake, the charges were dropped. Williams, who had been in police custody for a total of 30 hours, was released on bail.

Although his case was dismissed two weeks after he was arrested, it was dismissed “without prejudice,” meaning he could later be charged again, according to the New York Times.

On Wednesday, the American Civil Liberties Union (ACLU) of Michigan filed a complaint with the city, calling for the dismissal of Williams’ case and for his information to be removed from Detroit’s criminal databases. Prosecutors have since agreed to delete his data.

The Detroit Police Department “should stop using facial recognition technology as an investigatory tool as the facts of Mr. Williams’s case prove both that the technology is flawed and that DPD investigators are not competent in making use of such technology,” Phil Mayor, an attorney at ACLU, wrote in the complaint.

Over the years, various studies have provided proof that facial recognition tools are often biased against minorities. A 2019 study by the National Institute of Standards and Technology found that though the technology works relatively well on white men, it provides less accurate results for other demographics, and experts have blamed this on a lack of diversity in the images used to develop the databases.

In a statement to NPR, the Detroit Police Department said after the Williams case, the department passed new rules. It said now only still photos, not security footage, can be used for facial recognition and it is now used only in the case of violent crimes.

“Facial recognition software is an investigative tool that is used to generate leads only. Additional investigative work, corroborating evidence and probable cause are required before an arrest can be made,” the police department said.

Months after this unfortunate incident, Williams said he has not talked to his neighbors about it as well as his colleagues.

“My mother doesn’t know about it. It’s not something I’m proud of,” Williams said. “It’s humiliating.”

In a Washington Post op-ed, Williams wrote: “I never thought I’d have to explain to my daughters why daddy got arrested. How does one explain to two little girls that a computer got it wrong, but the police listened to it anyway?”

He said he thinks about the damage his false arrest could have on his family, especially his daughters, who have started playing games where people were arrested, and have even accused him of stealing things from them.