Associated Incidents
Arrested on November 25, while driving in the suburbs of Atlanta (Georgia), Randal Reid probably did not anticipate the turn this case was going to take. And for good reason, this 29-year-old African American was arrested and then provisionally detained for six days for nothing. Indeed, the man was apprehended for thefts he had never committed, and in states where he had not visited. He was released from prison on December 1.
"Imagine living your life and somewhere far away, you are told that you have committed a crime. And you know you have never been there", confided Randal Reid in the New York Times. If such an error was made, it was because of the use of the intelligence, Clearview AI, by local police. If no mention of this tool appears in the official report, a source close to the file was able to confirm it to the American daily, which returns at length to this police error. To help with its investigations, the sheriff's office signed a contract with Clearview AI in 2019, for $25,000 per year.
The racist prejudices of artificial intelligence
In fact, local police relied on CCTV footage from the town of Metairie, Louisiana, showing a man stealing Chanel and Louis Vuitton bags from a store for $10,000. These images were then submitted to Clearview AI, which identified Randal Reid as responsible for the mischief. This technology offers a search engine based on faces, using billions of photographs from the Internet.
She has been especially in the sights of politicians for a few months, being accused of racism by confusing and wrongly incriminating black men. [In 2019, American researchers](https://www.washingtonpost.com/technology/2019/12/19/federal-study-confirms-racial-bias-many-facial-recognition-systems-casts-doubt-their -expanding-use/) demonstrated that Asian and black people were up to 100 times more likely to be misidentified by Clearview AI than white people. Middle-aged white men have the highest facial recognition accuracy rate. In early March, a proposal was tabled by Democrats in Congress [to ban its use](https://truthout.org/articles/warren-sanders-want-to-ban-racist-facial-recognition-tools-from -government-use/) by law enforcement and government.
This is not the first time that American elected officials have urged the federal police, including the FBI, to stop using the tool developed by Clearview All. "Facial recognition tools pose a serious threat to civil liberties and public privacy rights, and Clearview AI's product is particularly dangerous," three senators said in February 2022.