Facial Recognition Platform Misidentified Dozens of Politicians as Criminals
Technological advances have made life so much easier for millions of people, all over the globe. Whether it be a smart device in the home or the ability to order food from a mobile phone. Technology has changed the world and it's not just private citizens that are using it either, many governments and government organizations are also using new, modern ideas to help them. One example would be facial recognition. Facial recognition technology is being used by many police departments to assist in identifying criminals and consequently making arrests. However, there has been some considerable controversy in recent months surrounding the accuracy of such programs.
Facial recognition software is by no means perfect. Still, in its early days, there were a few teething problems to be expected but repeatedly making blatant errors is a bit heavy-handed. The AWS facial recognition platform powered by Amazon is still misidentifying people at an alarming rate, even as it’s being used by police departments to arrest and prosecute. Paul Bischoff, a consumer privacy expert with Comparitech, found that the Amazon platform incorrectly misidentified more than 100 photos of US and UK politicians as criminals.
Amazon’s cloud-based facial recognition platform, Rekognition, that was first launched in 2016, has been sold and used by a number of United States government agencies, including ICE and Orlando police, as well as a number of private entities. When comparing images of 1,959 US and UK lawmakers to subjects in an arrest database, it was found that Rekognition misidentified at an average of 32 members of Congress. Moreover, the platform was also found to be racially biased, misidentifying non-white people at a higher rate than white people.
Not only is this unacceptable in itself, but mistakes like this have real-world impacts on peoples’ lives when police are blindly following the software. One, rather disturbing example, would be when Detroit citizen, Robert Julian-Borchak Williams, was violently arrested and dragged into custody after a facial recognition system falsely matched his photo with security footage of a shoplifter.
The controversial nature of this incident has sparked the government to propose legislation that would indefinitely ban the use of facial recognition technology by law enforcement nationwide. Although tech giants such as Amazon, Microsoft, and IBM, have previously sold facial recognition to law enforcement and governments alike, all three have placed a moratorium on their platforms preventing the use of the technology to look for criminals.