HomeTechnologyWrongly accused of facial recognition, an African-American spends a week in prison

Wrongly accused of facial recognition, an African-American spends a week in prison

A 28-year-old black man was the victim of a facial recognition tool in the state of Louisiana. A case that illustrates the criticism of the racial biases of the algorithms.

On November 25, Randal Reid was driving with his mother in Georgia when he was arrested by police based on a warrant issued by Jefferson Parish County, Louisiana. He was immediately taken into custody. Except Randal Reid never went to Louisiana. Accused (incorrectly) of robbery, he was flagged as a suspect by a facial recognition system, reports the US site Ars Technica. He was only released from the DeKalb County, Georgia, jail on December 1 after police admitted to his mistake.

Blind trust in technology

As the local outlet Nola points out, the local police relied on CCTV footage from the city of Metairie, Louisiana. They date from the summer of 2022 and show a man stealing Chanel and Louis Vuitton bags from a store, for an estimated loot of $10,000. Images sent to a facial recognition system, which wrongly identified Randal Reid as the perpetrator.

To the American media, Tommy Calogero, Randal Reid’s lawyer, regrets that local authorities have given full confidence in the technology, without even carrying out elementary checks or noting noticeable physical differences, starting with a twenty-pound weight difference.

The name of the facial recognition tool used by the police has not been made public. Among the market leaders, the company Clearview, recently condemned by the Cnil for having illegally used the photos of French internet users to sell their services. To work, these systems compare video surveillance images with the millions of photos available on the Internet, for example, on social networks.

The skin color of the victim of this misjudgment is also not insignificant: whether it’s the ClearView tool or Amazon, facial recognition is regularly accused of numerous ethnic biases, due to which it works less reliably on darker skin tones. .

Author: Raphael Grably
Source: BFM TV

Stay Connected
16,985FansLike
2,458FollowersFollow
61,453SubscribersSubscribe
Must Read
Related News

LEAVE A REPLY

Please enter your comment!
Please enter your name here