It’s like a nightmare come true.
One of those scenarios that we often speculate about here on this blog actually happened to one Detroit woman who, after facial recognition software mistakenly identified her as a violent criminal, was arrested and held in custody while pregnant with twins.
It’s like a tale ripped straight out of science fiction with more than a touch of horror. And it has led to a lawsuit as Detroit resident Porcha Woodruff seeks to make sure it not only doesn’t happen again but that the public is aware of just how quickly things can go sideways when you rely upon technology to do a human eye’s job.
She was arrested as she was taking two of her children to school and initially thought the six squad cars were some kind of joke, particularly since she states that she was visibly pregnant, NBC News reports.
How was she mistakenly identified as a carjacker and robber? A victim of the crime for which Porcha Woodruff was wrongly accused identified her picture from a lineup generated by a computer using facial recognition technology. There are some other concerning elements to the story, though, that might limit the facial recognition technology’s fault here.
NBC News reports that a phone stolen from another victim of the assault was returned to the gas station where the crime took place. Analysis of the video showed a female and this was used as the basis for the warrant for Porcha Woodruff’s arrest which means that quite a few leaps of logic took place for all of this to happen.
The lawsuit details how the detective on the case observed the footage from the gas station and made no mention of the suspect being pregnant. Woodruff pressed officers on this fact, asking if the suspect they were looking for was pregnant, but she was taken into custody anyway and experienced complications with her pregnancy because of that, the lawsuit alleges.
Any thoughts you might have on facial recognition technology, artificial intelligence, and modern surveillance are welcome in the comments.
Check out some of our other headlines at this link.