Aug. 8, 2019, 3:51 p.m.
Facial recognition is currently enjoying a very bad name for fear of a surveillance state. Mass facial recognition indeed means governments can potentially know where everyone is, all the time. Facial recognition answers the question “who are you?” by comparing your biometrical facial features with a neural network, potentially created by a machine learning algorithm that have created hashes from every face on Earth. Facebook has most of our faces on file.
There is, however, a flipside, the answer to the following questions, “Are you who you say you are?” or “Are you allowed through this door?” or “Should I let you transfer one million bitcoin from one account to another?” or “Must I give you access this top secret data?” or “Is it you who is really signing this document?”
Facial Authorization with AI is Secure
Facial authentication answers those questions and addresses entirely different use cases. Authentication actually increases our individual security rather than decreases it. Banks, airports, data centers, hospitals, to name a few, are places where authentication raises the quality of interactions. Would you want a surgical team to know that your face didn’t match a kidney patient’s facial dataset before they took your kidney out instead of theirs?
Digital Identity Authentication (DIA) is being built around biometrics already, and simply comparing the photo on an identity card with a person’s face at the door of a club or office or airport would raise the bar and avoid the human error caused by laziness, or a close match on the part of a person trying to gain entry to a building with a flash drive of viruses.
However, it’s facial authorization that we really want. The rash of so called SIM-swapping hacks that has led to massive bitcoin thefts could be stopped with facial authorization of transactions. Bitcoin exchanges like Coinbase require new accounts to be opened with a government issued photo ID. Comparing the face of a transactor with their photo ID is unspoofable with liveliness detection from Chooch.
Face ID Match Authorization Can Stop Bitcoin Theft
Likewise, adding biometrics such as facial authentication at key points where identity checks such as passwords, badges, PIN numbers, and wristbands are being used makes perfect sense. Triple authentication authorizes someone to take the next step. Companies like Docusign could and should ask a signatory to provide an ID, and require a face that matches the ID be present to officially sign a document
Facial authentication is already here, if imperfect. Half the population of the planet is using mobile devices as ID to access data from and through their phones. The imperfection of facial authentication is simply not knowing the answer to questions such as “Are you trying to spoof me?” or “Is the face in front of a camera two dimensional?” or “Is that face alive?”
At Chooch, our visual AI technology does “know” the answer to those “aliveness” questions. When our neural network perceptions encounter a face, we can ensure that the face is 3D instead of 2D. That’s part of the logic behind our endorsement of facial authorization as a key feature of our future collective security.
Privacy is less is a concern when technology is used simply to prove that person A is indeed person A. When machine learning is applied to a dataset of images of a face, a hash is created. The number that represents the face can rendered completely unreadable to any other system. Secured thus, this lock for which a face is a key cannot be reverse engineered or altered, so only the original face, alive and in 3D, can generate an unlocking.
Facial authorization means that you will be sure your housecleaner is unlocking your front door when you are not home. It means that trillions of financial transactions will not be unlocked by spoofable individual identities. It means only one face can unlock a kidney removal at a particular place and time, say 3PM on a Tuesday. And it means no one can forge your signature unless they have your face, because Chooch can even tell twins apart.