This you?
Imagine pulling into the driveway at your home, and police arrest you solely based on the fact that a computer said you looked like a criminal. This nightmare was a real experience for Robert Julian- Borachak Williams, a black man who was falsely arrested after a facial recognition system identified him as a man who was wanted for non-violent robbery. After being held in jail for 30 hours the police recognized that the photo actually did not match Williams at all, and his case was dismissed ‘without prejudice’(Porter). This means this specific charge is dropped, but he could be charged again with the same crime(https://www.robertmhelfend.com/criminal-defense/dismissed-without-prejudice-mean/).
Facial Recognition is a computer operating system that is capable of detecting, tracking, identifying or verifying human faces from an image or video from a digital camera (https://www.sciencedirect.com/science/article/pii/S2590005619300141). Technology companies sell their facial recognition programs to the government which is how the police are able to use it to identify criminals or people they believe to be criminals. Facial recognition has been proven time and time again to show racial bias when analyzing certain groups of people, yet it is still being utilized to uphold the law in the U.S. This dangerous algorithm would classify facial recognition as a WMD or a “Weapon of Math Destruction.”
WMD is a term coined by data scientist Cathy O’Neil. She defines a WMD as a machine that allows “math to let assumptions go untested and unquestioned” (O’Neil). So when the facial recognition system identified Williams as being a wanted robber, the police did not even question if it was true. They trusted the faulty algorithm and automatically arrested a man without double checking the photo captured by the machine.
There are plenty of other cases where facial recognition’s bias build has falsely identified people in society. Some facial recognition experts even believe that there should be a delay on facial recognition use because it is so flawed. However, law enforcement, an already controversial organization, is still able to utilize this tool to put people in jail.
As a solution, facial recognition is only utilized by the police when investigating violent crimes, but this is not enough. Today, data scientist are not motivated to make quality products with minimal to no bias because the incentive is selling it to make the most money. So does a system such as facial recognition need more time to be adjusted and regulated or does it need to be destroyed altogether?
References
Bah, S., & Ming, F. (2019, December 26). An improved face recognition algorithm and its application in attendance management system. Retrieved February 12, 2021, from https://www.sciencedirect.com/science/article/pii/S2590005619300141
O’Neil, C. (2017). Weapons of math destruction: How big data increases inequality and threatens democracy [Introduction]. In 1192096331 891157873 C. O’Neil (Author), Weapons of math destruction: How big data increases inequality and threatens democracy (pp. 9–18). Great Britain: Penguin Books.
Porter, J. (2020, June 24). A black man was wrongfully arrested because of facial recognition. Retrieved February 12, 2021, from https://www.theverge.com/2020/6/24/21301759/facial-recognition-detroit-police-wrongful-arrest-robert-williams-artificial-intelligence
