A black man has filed a lawsuit against policeover facial recognition technology that landed him in prison, joining the ranks of other African Americans who have also been victims 's racial bias technology.
Facial recognition technology's racial bias against non-white skin landed Nijeer Parks in jail for 10 days in 2019 after he was wrongly identified as suspect theft.
Parks says in his lawsuit that it didn't matter that he had never been to the scene of the crime, a Hampton Inn in Woodbridge, New Jersey. The technology identified him as a suspect, and apparently that was enough for the police. A warrant was issued, and Parks tried to explain that they had caught the wrong man.
“I had no idea what was going on,” Parks told NJ Advance Media. “I had never been to Woodbridge, I didn’t even know exactly where it was.”
But according to Parks, no one heard of it. Once he arrived at the police station, he was handcuffed and taken to jail for ten days. The charges against him were later dropped, and prosecutors admitted that the only evidence against Parks came from the Clearview AI facial recognition technology.
“I had a background, but I’d been home since 2016 and I hadn’t had any problems,” Parks said. “The whole thing really scared me. I’m trying to do the right thing in my life.”
Now Parks is suing the police department for arresting him based solely on faulty technology facial recognition The worrying thing is that he's not the only one to whom this has happened.
Robert Julian-Borchak Williams was also arrested earlier this year in Detroit for stealing watches in 2018 after he was also misidentified by facial recognition technology. This time, it was DataWorks Plus recognition software used by Michigan police.
According to Business Insider, Williams was arrested in front of his wife and children and spent the night at the Detroit Detention Center, before being taken to an interrogation room where he was shown the video of the robbery. That's when officers realized they had made a mistake and that it was someone else.
Racial bias in facial recognition software
According to a Harvard study , about half of American adults are included in police facial recognition databases without their knowledge or consent. This is a major privacy issue for all Americans.
However, African Americans are in an even worse position, as they are more likely to be wronged by the technology itself. According to Harvard’s Alex Najibi, overall accuracy rates for facial recognition technology hover around 90%, but error rates vary across different demographics. “Lower accuracy was consistently found among women, people of color, and people aged 18-30,” he wrote.
The idea that facial recognition software is racist has been hotly debated in recent years, and other research has found that facial recognition technology is less accurate for Americans of color, especially women.
Removing the wrong technology
Last November, the Los Angeles Police Department banned the facial recognition platform Clearview AI, following pressure from the American Civil Liberties Union (ACLU).
“ Clearview AI records information, without our knowledge, and uses everything from casual selfies to photos from birthday parties, graduations, weddings, and more, ” wrote ACLU attorney Nathan Freed Wessler .
“Without the public's knowledge, this company has offered this massive database to private companies, police, and federal agencies, allowing them to secretly track and target anyone they want using facial recognition technology.“.
Additionally, in recent months tech giants such as Microsoft , Amazon , and IBM have pledged to stop selling facial recognition to police agencies, to protect user privacy .
Clearview CEO Hon Ton-Thatdefended his company's product, emphasizing that it is used by law enforcement agencies to detect crimes and protect people.
Source: Threatpost
