Face-Reading Tech Leads To False Arrest Of Man Who’d Never Been To Crime Scene

Facial recognition is often at the center of discourse for its innate bias, in that it is more likely to misidentify a person of color than a white person. And to prove this point, a Black Georgia man was imprisoned by police after they used facial recognition and identified him as a criminal.
 
Police had taken Randall Reid, 28, into custody in November in DeKalb County, Georgia, on the grounds of allegedly stealing luxury purses from a consignment shop along with three other accomplices. These thefts reportedly took place in Jefferson Parish and Baton Rouge, Louisiana.
 
Police in Baton Rouge used identification from the Jefferson Parish Sheriff’s Office to obtain an arrest warrant.
 
According to a report from The Times-Picayune (via The New Orleans Advocate), Reid says that he was told the thefts had happened in Louisiana, which he has never been to before, nor has he ever committed a crime.
 
His attorney, Tommy Calogero, noticed a mole on Reid’s face that differed from the actual suspect and a difference of 40 pounds in weight between them. With this evidence, Calogero prompted the Jefferson department to review his case and rescind the warrant for his arrest. 

Reid was initially apprehended on November 25 and released on December 1.
 
The New Orleans police clarify that facial recognition as mass surveillance is only used to generate leads. Officers must get approval from department officials and the Lousiana State Analytic and Fusion Exchange in Baton Rouge before making an arrest.
 
That being said, there are organizations out there that are actively working on removing other prejudices such as gender bias from AI programs. 
 
 
 
[via Associated Press News and The Times-Picayune/The New Orleans Advocate, Photo 137534981 © Viktoriia Kotova | Dreamstime.com]