Ultimately, facial authentication will be a prevalent method in anything that requires more security but eventually trickles down to every industry, including replacing keys and card readers to become the de facto access-control-authentication method. Facial authentication shouldn't be confused with facial recognition. The latter is prone to misuse, as governments, corporations, and other entities leverage public data (as well as data collected without consent) to identify or track people without their awareness. This isn't what Apple's FaceID does, and it's not what access-control solutions do either. Rather, authentication matches specifically enrolled, consenting people to their own identity without plugging into significant databases.

There's a lot of misinformation in this comment section, and it probably doesn't help that the linked article isn't particularly well-written. I'll try to correct a few misconceptions as I understand them. I am a machine learning engineer although I don't work in CV so my knowledge isn't highly specific there. Fundamentally, facial recognition (FR) is a function which takes as input an image of a face (or faces) and outputs an identity. A closely related technology is used for authentication, like on newer phones.

The field of computer vision seeks to help computers "see" images and videos at a high level, in the same way that humans can--but within this domain, there are many different subfields and applications. Facial authentication is a task in computer vision that seeks to verify a person's identity by using their facial characteristics and features as biometric identifiers. Facial authentication is similar to, yet distinct from, facial recognition. The term "facial recognition" describes any system that can identify individual people via their facial features, often using surveillance videos as input. However, facial authentication is strictly opt-in: individuals knowingly present their face to the system, usually for security purposes.

When bias becomes embedded in machine learning models, it can have an adverse impact on our daily lives. It's exhibited in the form of exclusion, such as certain groups being denied loans or not being able to use the technology. As AI continues to become more a part of our lives, the risks from bias only grow larger. In the context of facial recognition, demographic traits such as race, age, gender, socioeconomic factors, and even the quality of the camera/device can impact software's ability to compare one face to a database of faces. In these types of surveillance, the quality and robustness of the underlying database is what can fuel bias in the AI models.

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?"

When a person gains access to a secure building, sensitive data, or vast sums of corporate finance via facial authentication, how do you know they are who they say they are - for sure? Masks from The Real Face Japan. Yes, Chooch AI has a neural network model and learns the 512 biometric features of every face it is trained to learn. Yes, when an image of a face is sent via API to Chooch AI, we have a higher than 99% accuracy in facial authentication. We know whether that face has been learned by Chooch or not, but what happens if the person trying to gain access has a very, very good mask?