It's almost inevitable for facial recognition to be used by the authorities. It can help them identify perpetrators through camera footage more easily. The problem is that the system doesn't always get right, and that can lead to undeserved consequences on a person's life.

Detroit Police's New Facial Recognition Rules

It's unfortunate how there have been many wrongful arrests in the past and facial recognition has just made the situation worse. Acknowledging how it is not ready to be the sole basis for arrests, new rules are set by the Detroit Police Department.

What's worse is that the change was prompted by yet another wrongful arrest of a Black individual, Robert Williams. He was arrested in 2020 after a shoplifter's face was "identified," which was then used to construct a photo lineup.

Without further evidence, Williams was arrested in his home, which the victim of false recognition said "completely upended" his life. The court has finally reached a settlement with the wrongfully accused and has banned the technology, as per Engadget. 

However, it won't be a permanent ban since the agreement will only restrict the use of facial recognition technology for leads for four years. The American Civil Rights Union (ACLU) called the restriction a "groundbreaking settlement agreement."

The summary of the agreement states that an "FRT lead, combined with a lineup identification, may never be a sufficient basis for seeking an arrest warrant," and that there must be further independent and reliable evidence linking a suspect to the crime.

Detroit Police will also have to undergo training and learn about the technology's tendency to be inaccurate and biased, especially with people of color. All cases that used facial recognition technology since 2017 will be audited.

Read Also: Facial Recognition Results in Pregnant Woman Being Wrongfully Arrested

Bias in Facial Recognition Technology

This is not the first time that a Black person has been mistakenly incriminated because of facial recognition, and studies have shown that the identifying technology is more likely to misidentify people of color leading to wrongful arrests.

Another woman from Detroit who was eight months pregnant at the time of her arrest has already filed a lawsuit against the police for the same incident. The victim, Porcha Woodruff was arrested by six police officers accusing her of robbery and carjacking, as per NBC News. 

The court document states that Woodruff only later discovered that she was implicated as a suspect, also through a photo lineup shown to the victims of the robbery and carjacking. To make matters worse, Woodruff was only released on a $100,000 personal bond.

The authorities on the case reportedly used an eight-year-old photo of Woodruff when she was scanned even though they had access to her driver's license. They even refused to check if the suspect in the warrant was pregnant like Woodruff was, which could've cleared everything up easily.

In a different case, Michael Oliver was one of the three people who were accused of a crime he did not commit. His arrest led to severe consequences as well, such as his losing his job in 2019. It took three misidentifications before the police finally made changes in how facial recognition was used.

Related: Detroit Compensates Wrongly Accused Man with $300,000; Updates Facial Recognition Technology