Following a settlement with Robert Williams, a Black man wrongly arrested in 2020 due to a mistaken facial recognition match, the Detroit Police Department will now adopt stricter rules for using facial recognition technology.
These rules, in place for four years, aim to prevent similar mistakes. Police cannot arrest someone just based on facial recognition or use it alone to create lineups. These changes are seen as the toughest yet in policing across the country, according to the ACLU.
Williams’ ordeal began when his expired driver’s license photo was incorrectly flagged by facial recognition software as matching an alleged shoplifter. Subsequently, police used this flawed identification to assemble a photo lineup and arrested Williams at his home in front of his family, an event he described as profoundly disruptive to his life.
Notably, similar incidents of wrongful arrests based on facial recognition have disproportionately affected Black individuals, a trend highlighted by the ACLU.
The settlement mandates that facial recognition alone cannot serve as the primary basis for seeking arrest warrants. Police must corroborate such leads with additional independent and reliable evidence linking a suspect to a crime.
Moreover, Detroit PD officers will undergo training addressing the technology’s racial biases, aiming to mitigate its historically higher rates of misidentification of people of color. Additionally, an audit will review all cases since 2017 where facial recognition was pivotal in obtaining arrest warrants, ensuring compliance with the new standards.
In an op-ed for TIME, Williams expressed that the agreement signifies a critical shift away from overreliance on facial recognition at the expense of traditional investigative methods. This settlement reflects growing awareness and concern over the discriminatory impacts of facial recognition technology, particularly on marginalized communities.
As Detroit moves forward with these reforms, the case of Robert Williams stands as a pivotal example, prompting policy changes aimed at safeguarding civil liberties while addressing systemic biases in law enforcement technology.
