Clear dangers of photo tech used by police

Few enough American women while eight months pregnant and already taking care of two young children could muster the energy to engage in a complicated armed robbery and a carjacking.

Yet that’s exactly what Detroit police accused Porcha Woodruff of last February when six officers showed up at her home and arrested her for the crimes in front of her two daughters, even after she gestured to her pregnant belly and scoffed, “Are you kidding?”

The basis for their arrest was what is rapidly becoming one of the clearest present dangers to freedom in this country under the guise of crime fighting: Wildly false results that come from improper use of automated facial recognition technology in which law enforcement utilizes the technology to match photos from a database.

As Kashmir Hill reported this week in The New York Times, Woodruff, a nursing school student with no record other than a minor citation for driving with an expired license eight years ago, was nonetheless hauled away and charged in court for the robbery and carjacking. After a full day in the Detroit lockup, Woodruff was released on a $100,000 personal bond. “In an interview, she said she went straight to the hospital where she was diagnosed with dehydration and given two bags of intravenous fluids. A month later, the Wayne County prosecutor dismissed the case against her,” the Times reports.

The case was dismissed because Woodruff had nothing to do with the crime, as more old-fashioned police work surely would have shown at the outset.

Each mistake involving facial technology — many of which involve wrongly accused suspects who happen to be Black — surely is mistaken in its own way. In this case, Detroit police used the equipment from facial recognition company DataWorks Plus “to run unknown faces against a database of criminal mug shots; the system returns matches ranked by their likelihood of being the same person. A human analyst is ultimately responsible for deciding if any of the matches are a potential suspect. The police report said the crime analyst gave the investigator Ms. Woodruff’s name based on a match to a 2015 mug shot.”

That mug shot is blurry and could depict the face of thousands of women in the Detroit area. What police did not turn to was the easily accessible photo from Woodruff’s current driver’s license, which is clear and crisp. Instead, they showed the mug to the victim of a robbery and carjacking that occurred at a local gas station and “asked the victim to look at the mug shots of six Black women, commonly called a ‘six-pack photo lineup.’ Ms. Woodruff’s photo was among them. He identified Ms. Woodruff as the woman he had been with. That was the basis for her arrest, according to the police report.”

Woodruff, the third person in Detroit to recently report being the victim of mistaken identity owing to faulty facial recognition searches, has understandably filed suit. An attorney for her case notes that the Detroit Police Department “is an agency that has every reason to know of the risks that using face recognition carries,” said Clare Garvie, an expert on the technology at the National Association of Criminal Defense Lawyers. “And it’s happening anyway.”

We have problems — clear constitutional ones — with the rampant and rapid deployment of intrusive photographic technology at all levels in American society by law enforcement. Even when it sometimes fingers the right suspect, do we really want each moment of our public lives to be on candid camera?

But since the current technology is so faulty, and police are so ill-trained in its use, it’s time to call a halt to the employment of automated facial recognition systems. “It is circular and dangerous,” a psychologist who has studied them said. “You’ve got a very powerful tool that, if it searches enough faces, will always yield people who look like the person on the surveillance image.”