Facial Recognition Software Can Lead to Injustice
In February of 2019, Nijeer Parks, 31-years-old at the time, was sent to a corrections center in New Jersey due to a false facial recognition match placing him at a grocery store at the time of a crime.
Mr Parks was accused of shoplifting candy, hitting a police officer’s vehicle, and nearly driving into a law enforcement officer upon escape. Mr Parks’s identity had been weeded out through facial recognition software that declared he was present during the crimes, despite Mr Parks claiming his innocence.
In fact, Mr Parks was 30 miles away from where the crimes had occurred, yet due to this software and how much Mr Parks resembled the fraudulent license photo of the man responsible, police issued a warrant for his arrest.
Mr Parks spent 10 days in jail while also paying to defend himself for the crimes that he did not commit. In November of 2019, however, the case was dismissed due to a shortage of evidence pinning Mr Parks to the alleged crimes.
Now Mr Parks is suing law enforcement for their horrific mistake that could have ruined his life. Mr Parks is suing the City of Woodbridge police department on charges of false arrest, false imprisonment and violation of his civil rights.
These instances of flawed facial recognition raise the question of whether there have been more wrongful incarcerations in the past, and if other innocent citizens are still sitting in jail for crimes that they did not commit. This software predominately harms people of color, despite law enforcement continuing to defend it.
Clearview AI is a facial recognition tool that law enforcement use to enhance their cases, and to find those who could have potentially committed crimes. However this software is not immune to mistakes, as there have been two previously falsely arrested citizens before Mr Parks who had warrants for their arrest due to faulty facial recognition matches.
All persons known to have been wrongfully accused in the past have been Black men, as the software does not work appropriately with Black and Asian facial structures. In Mr Parks’s case, police used a license photo residing in a government database, that of which this AI system does not have proper access. The final analysis on how Mr Parks could have been the sole suspect raises many questions on the development of law enforcement tools.
Mr Parks had previously been arrested due to selling drugs, but was then released in 2016. However, due to his past convictions, the faulty facial recognition match could have locked him away in jail for a longer sentence due to it being his third felony. Mr Parks claimed that when the prosecutor offered a plea deal in his case, he nearly took it out of fear, despite being completely innocent of the crime.
This is just one example of law enforcement potentially incarcerating yet another citizen for crimes that they did not commit. Laziness and dependence on software that does not have a high degree of accuracy in identifying determining criminals could jeopardize the futures of many other citizens of color.
Our law firm supports the development of standards to make sure that questionable applications of technology do not deprive citizens of freedom.