In a legal settlement, the Detroit Police Department has agreed to cease using facial recognition results as the sole basis for arrests. This decision comes after the wrongful arrest of Robert Williams, a man of colour, who spent 30 hours in jail due to a flawed identification process.
The incident that led to this pivotal change occurred in 2020 when Williams was wrongfully arrested based on facial recognition technology that misidentified him as a shoplifter. The arrest was linked to a 2018 shoplifting case captured by surveillance cameras at a downtown Detroit watch store. About 15 months later, the footage was analyzed, resulting in Williams being wrongly identified.
In the 2018 incident, an unknown individual stole five watches, and the grainy CCTV footage was subsequently processed by the police using a facial recognition system. This system generated a list of 243 potential matches, including Williams’s old driver’s license photo. Despite being ninth on the list, Williams was incorrectly singled out as the best match, leading to his inclusion in a photo lineup presented to the store’s security contractor.
The security contractor’s confirmation resulted in a warrant for Williams’ arrest. He was detained, spent a night in jail, and underwent fingerprinting and DNA sampling. Charged with retail fraud, Williams had to hire a lawyer. It was later proven that Williams was at his workplace during the shoplifting incident, leading to the charges being dropped.
Outraged by his wrongful arrest, Williams sued the Detroit Police Department in 2021 to end the use of the faulty technology that led to his ordeal. Last week, the department settled with Williams, agreeing to pay $300,000 in damages and no longer relying solely on facial recognition for arrests. Additionally, the police department committed not to conduct lineups based on facial recognition without independent, reliable evidence linking a suspect to a crime.
The settlement also requires the police to disclose the flaws of facial recognition technology and the other potential suspects identified by the system. This move aims to increase transparency and prevent future wrongful arrests.
Facial recognition systems use biometrics to identify individuals by comparing facial features from videos or photos against a database. While widely used for security in stores, public areas, and schools, these systems have frequently misidentified innocent people, leading to wrongful arrests.
Williams’ case is one of many that highlight technology’s shortcomings. The American Civil Liberties Union (ACLU) has long advocated for stricter rules governing police use of facial recognition.
“This settlement moves the Detroit Police Department from being the best-documented misuser of facial recognition technology into a national leader in having guardrails in its use,” said Phil Mayor, an ACLU of Michigan lawyer. The Detroit Police Department expressed its commitment to using facial recognition responsibly for public safety, boasting “the strongest policy in the nation now.”