Researchers at the AI firm Kneron were able to show that the facial recognition software used ubiquitously these days in banks, airports and checkpoints at various international borders, can be vulnerable to subterfuge as they were able to deceive those systems using sophisticated masks that they have developed, as reported to Fortune.
Moreover, they also mentioned that literally anyone who has the means to develop such a mask of high quality would be able to follow suit and be able to defeat the systems that employ facial recognition software, throwing caution to the winds for the many companies that jump into the bandwagon these days.
The researchers’ exploits are note-worthy in that they were able to fool several systems with ease, particularly a self-boarding terminal at Schiphol Airport in Netherlands where they were able to deceive the system just by using a picture of a face on a phone screen.
They noted that some systems were more easier to manipulate than others, like the airport example above, nonetheless they were able to even fool the payment systems by Chinese tech giants Alibaba and WeChat by scanning the face of a person registered in the system, whom they chose as their victim and wearing that person’s mask.
According to Fortune, they did mention the difficulty others would face when trying to replicate their product as theirs is a novel mask from Japan.
Kneron’s CEO, Albert Liu told Fortune that with many inefficient and sub-par systems doing their rounds as a genuine and trustworthy facial recognition AI, the consumers are at a grave risk which cannot be ignored.
His conclusion evokes a deep sense of foreboding as it seems to imply that though there are measures available to the companies to upgrade their security, thereby precluding such attempts, they are unwilling to do that as they do not put security of the consumers first.
They also added that Apple’s iPhone X’s FaceID feature was impervious to their chicanery.