Detroit police are being used by a woman who claims she is the victim of wrongful arrest prompted by a facial recognition match. Detroit Police say facial recognition was not used in the case.
The lawsuit highlights the tension between concerns that law enforcement use of facial recognition will increase false arrests of people with dark skin and concerns that without it, crime will spiral out of control.
A pair of researchers affiliated with Georgia State University write in an editorial for The Washington Post that homicide rates have fallen significantly over time in places in the U.S. where police use facial recognition, relative to elsewhere.
In 2000, the murder rate per 100,000 people was 6.98. Statistics comparing jurisdictions where facial recognition was used with everywhere else show lower murder rates in 2003, 2007, 2016 and 2020. In the last instance, there were 9.81 murders per 100,000 people in places where the technology is not used by law enforcement, and 8.09 where it is.
The researchers argue that facial recognition both speeds up arrests and acts as a deterrent, demonstrating the likelihood of getting caught even in cases where it is not obvious to potential offenders that it is in use.
An investigation by the Post last year showed that U.S. police have a long way to go to meet public expectations for transparency in their use of facial recognition.
Not again . . .
LaDonna Crutchfield alleges that her arrest for attempted murder in January, 2024 was prompted by a false positive match made by a law enforcement facial recognition system. Her attorney says police married out “no investigation” before arresting Crutchfield in front of her children.
Detroit Police Department Assistant Police Chief Charles Fitzgerald says Crutchfield was identified as a suspect through video footage showing a beige Ford Fusion and a partial license plate. The vehicle was traced to the house where Crutchfield was arrested.
“She is alleging that we use facial recognition, and we did not,” Fitzgerald told NPR affiliate Michigan Public.
Crutchfield was shown a photo of a suspect from the crime scene, and claims she was told she had been matched with facial recognition. She was forced to submit fingerprint and DNA biometrics, according to Detroit Metro Times, and released later that evening.
Michigan police adopted a new governance policy for facial recognition use last year following a pair of wrongful arrests of Black people in Detroit.
Multiple media reports claim that facial recognition was used, though no corroboration for the claim is offered, and the Metro Times mistakenly states that all people arrested in the U.S. through false matches by facial recognition systems have been Black. Still, the real number, six out of seven, does little to dispel accusations that the technology is exacerbating police biases.