There are two main reasons opponents of proliferation of facial recognition and biometric data collection cite when they speak out against the trend.
One is invasion of privacy of each person subjected to it on a grand scale, and the other the inadequacy of the technology itself, given the task it is supposed to perform: mass surveillance that should produce conclusions, some of which can lead to serious consequences.
Now those who focus on the harm that comes from facial surveillance being error-prone have their first case they can cite as proof of the claim.
Namely, Porcha Woodruff is suing the city of Detroit alleging false arrest based on inaccurate facial recognition data.
The local police managed to get an arrest warrant based on that data and came to Woodruff’s home to execute it, in front of her children. Woodruff was also eight months pregnant at the time, but that didn’t dissuade the officers that she was in fact a suspect in a robbery and carjacking case.
We obtained a copy of the lawsuit for you here.
The warrant was based on a photo match obtained through facial recognition.
The victim also told the police that he first visited with a woman, then went to another location where the robbery and carjacking happened (perpetrated by a man, a co-conspirator), and that his phone then showed up at a gas station several days later.
It was from a video taken from the phone that detective LaShauntia Oliver – who is now a plaintiff in Woodruff’s case – extracted the images that were supposed to produce an accurate facial recognition match.
Woodruff’s photo was in the police database to check against because she had been arrested and in a line-up – but this was eight years ago, the lawsuit says.
Her visible pregnancy seems to be the key of the allegations against the police and the city, since it was not mentioned in what is referred to as a detailed report Oliver submitted about the contents of the video found on the phone.
And the police apparently refused to confirm that the woman who was behind the crime was pregnant.
