Artificial intelligence (AI) may not always be our friend. Ever-increasing surveillance in public areas is creeping a lot of us out. A large portion of us would rather fly under the radar, thanks. And now, if you are prepared to wear wildly hideous clothing, there’s a good chance you will fool the facial recognition software that is all around us.
An Italian startup called Cap_able has designed a line of knitted garments that they claim will protect you from facial recognition software. And this is without the need to cover your face.
The clothing features a pattern known as an ‘adversarial patch’ developed by AI algorithms to confuse facial recognition software.
These garments are made to resemble animals like giraffes and zebras. According to Cap_able co-founder Rachele Didero, most cameras are designed to filter out animal information, whether that be dogs caught on street cameras or kangaroos that eat your lawn at night and fail to show up on your doorbell camera.
AI face recognition software in public places
As facial recognition technology becomes more pervasive, concerns about personal privacy have grown.
Governments and corporations around the world are already using AI technology to track and identify individuals. While this is good for catching criminals on the hustle, many average citizens feel that it might not be worth the price of admission.
According to documents posted by HWL Lawyers, in Australia, face recognition AI is used by banks and phone companies for identity verification. It is also used on our borders by immigration authorities and law enforcement agencies in Australia.
In South Australia, the South Australian Police have an AI called “NEC NeoFace system” in the Adelaide CBD, and have done so since 2019.
This kind of facial recognition technology can be used to identify someone by checking against a photo to see if it is the same person. So crims on the run have a harder time getting away.
AI confusion hit rate
Didero says the clothing works 60 to 90% of the time when tested with a commonly used algorithm in facial recognition software.
The clothing range that Cap_able makes was born from Didero’s Ph.D. at the University of Milan. The clothing, it says, can scramble the software for now.
But for how long? Companies and governments could easily add the clothing to their databases, classifying them as human. And with a huge price tag of €420 for a jumper, (AU$675) only wealthier people can afford them, so they aren’t likely to be in heavy circulation.
It seems it is expensive to be a dissident at this point in time. But in the future, it might be a very small price to pay.