Really good article in The New York Times about Robert Julian-Borchak Williams, who was falsely accused and arrested for robbing an upscale boutique in Detroit. He was accused of stealing $3.8K worth of watches. The smoking gun? A facial recognition algorithm used by law enforcement that turned out to be faulty.
The way the algorithm works is it scans possible suspects based on their driver’s license photos and Williams’ photo matched what the police thought was the suspect. Instead, it turns out Williams was innocent and that he wasn’t the man in the photos. As you can imagine, this was a very stressful experience for Williams, his wife, and his kids.
What adds to the controversy is the fact that Williams is an African-American and facial recognition technology has a tougher time accurately identifying African-Americans, Asians, and Latinos than Caucasians. It does seem like the case of Williams, many errors were made that led to his false arrest. It’s not strictly the facial recognition’s fault. More blame should go to the police for overly relying on the facial recognition and not checking for other clues.
Defenders of facial recognition technology will say it should be used as a tool to help police identify suspects as opposed to being a smoking gun. It seems like in this case the police overly relied on the technology and it backfired. The end result was the case being dropped, but “with prejudice.” Which means the case could be opened up against Williams in the future.
The A.C.L.U. of Michigan has since filed a complaint to the city of Detroit on behalf of Williams, asking for an absolute dismissal of the case, an apology, and the removal of his name from the cities criminal databases. It’s disappointing that more fighting had to be done after it was acknowledged that the facial recognition system failed to do its job.
Obviously, this is just another instance of a person of color being falsely accused of a crime that they didn’t commit and what’s worse is it looks like technology, a tool that is supposed to help prevent this sort of thing, can be used against them if not used properly.
(Photo: Sylvia Jarrus for The New York Times)