Zoom Image
Porcha Woodruff (on August 7)
Photo: Carlos Osorio / AP
Security authorities around the world hope that facial recognition software will be more successful in tracking down criminals. Critics, on the other hand, complain not only about data protection problems but also about the susceptibility of the programs to errors. What can happen when such a system spits out incorrect data was experienced by a woman in the northeastern United States.
The heavily pregnant woman had apparently been mistakenly arrested for a car robbery due to weaknesses in facial recognition software. Porcha Woodruff has gone to court and sued the city of Detroit for damages in an undisclosed amount.
According to media reports, the 32-year-old African-American woman is the first known case of a woman in the United States who was mistakenly detained on the basis of facial recognition software.
Woodruff, then eight months pregnant, had been arrested by half a dozen police officers on the morning of February 16 as she was preparing her two children for school. "She was perplexed by her visible pregnancy and thought it was a joke," the lawsuit says.
Victim of the robbery claims to have recognized Woodruff
Police in Detroit, Michigan, were investigating an armed car robbery involving a woman. She had images from a surveillance camera evaluated by facial recognition software, Woodruff appeared as a possible hit. The victim of the car robbery was then shown the photos of several women, including a police photo of Woodruff from 2015. The victim identified Woodruff as the perpetrator.
The 32-year-old was arrested and released on bail later today. The case was later dismissed due to lack of evidence.
The woman now argues in her civil lawsuit that the police should not have relied on software that is error-prone, especially in the identification of African Americans. According to experts, there is always confusion with such programs, especially among blacks.
jok/AFP