Skip to Main Content

Facial recognition tech is a potentially unreliable police tool

A Detroit man had to spend 30 hours in jail recently over a shoplifting charge he had nothing to do with.

Police arrested the man, who is African American, while he was at home with his family. The did so based on so-called facial recognition technology.

Police moved to arrest the man after reviewing footage from the store where someone took watches without paying for them.

While the image in the video was blurry, facial recognition technology allowed them to compare features to other images known to be that of the man whom they accused. In this case, though, police ultimately had to admit that the technology failed and had led them to arrest an innocent person.

There are serious concerns about the police’s reliance on this technology. For one, many times this technology uses photographs that the government requires.

For instance, driver’s license photos can be stored and then used in a facial recognition technology system for comparison to, say, fuzzy surveillance videos which show a crime in progress or other critical evidence.

Of course, people routinely get their photos taken to get a driver’s license without any inkling that it will eventually be used against them.

Moreover, as in the recent case, this technology can be unreliable and lead to false arrests and even false convictions. Some have said that the technology even can give credence to racial bias among police officers who are using it.

A person who is accused of a crime based on facial recognition technology may have legal options available to him or her.  For instance, he or she may want to consider asking the judge to disallow the evidence on various legal grounds.