Randal Quran Reid is among at least five Black plaintiffs who say they were misidentified by facial recognition technology and then wrongly arrested.
Randal Quran Reid was driving to his mother’s home the day after Thanksgiving last year when police pulled him over and arrested him on the side of a busy Georgia interstate.
He was wanted for crimes in Louisiana, they told him, before taking him to jail. Reid, who prefers to be identified as Quran, would spend the next several days locked up, trying to figure out how he could be a suspect in a state he says he had never visited.
A lawsuit filed this month blames the misuse of facial recognition technology by a sheriff’s detective in Jefferson Parish, Louisiana, for his ordeal.
“I was confused and I was angry because I didn’t know what was going on,” Quran told The Associated Press. “They couldn’t give me any information outside of, ‘You’ve got to wait for Louisiana to come take you,’ and there was no timeline on that.”
As they should
It’s been known for literally years that every AI based facial recognition algorithm has had a massive problem with bias in training data (more areas with PoC are policed, leading to more arrests of PoC, despite similar crime rates) as well as distinguishing dark colored faces from one another. Any application is automatically flawed and is not impartial in any way.