Facial recognition technology, while offering convenience and enhanced security in various applications, has sparked significant privacy concerns and debates over its potential for misuse. As reported by the ACLU and other privacy advocates, the widespread deployment of this technology raises critical questions about individual freedoms, data protection, and the risk of automated discrimination, particularly affecting marginalized communities.
Biometric technology has come a long way and with its progress came a very useful yet alarming innovation: AI facial recognition. It is exactly what its name implies— a technology that can scan a person's face, keep a record of their features, and use those records to identify that specific individual the next time the system comes in contact with their image or likeness12. It's similar to when you meet someone for the first time and you remember their face just by looking at them. Because their face has been added to your memory, you can easily recognize them when you cross paths again. Even when you meet other people or you're both in the same packed space like a party or a conference, you can still pick them out from the crowd because you know what they look like34.
Artificial intelligence can recognize faces just like humans, but our ways of doing so are different. Whereas people can simply discern one another with a single glance, AI can only determine and recall a person's appearance after a series of four steps. The process goes like this:
Face Detection: AI is trained to identify a face based on a clear image or video of an individual. Once either input is uploaded into the system, AI algorithms process the visual media right away until they find and separate the face from the background and other elements12.
Face Analysis: AI continues to scan the face with deep learning and Convolutional Neural Networks (CNNs) until it determines the facial features, accurately figuring out details such as eye distance, cheekbone contour, nose length, mouth position, and jaw shape34.
Face Capture: AI converts the visual information it has collected into digital data, which will be stored so the system can recollect the same face moving forward5.
Face Matching: The processed face will be compiled and contrasted in the AI's database of saved faces to differentiate it and establish its unique identity67.
Face recognition has become a vital component of a lot of industries, especially for security and privacy purposes. You may notice this as a feature in your smartphones, usually called "face unlock" where the phone scans your face and will only open once it recognizes you as its owner1. This deters other people from using your phone without your permission. The same system is integrated into banking and finance apps as well, so no one can access your accounts and touch your savings and investments except for you. Even some cars have this technology and won't let you inside unless you're authenticated2.
Identification is another major reason why AI facial recognition is important. In law enforcement, the police and military go through their database to pinpoint suspects, hunt criminals, and discover victims or missing persons3. Immigration does it too but for verifying travelers' identities and making sure that only those without a current criminal record or pending legal case are allowed to board an airplane or pass through a country's border4. Hospitals and healthcare facilities use facial recognition for identifying patients, while schools and companies use it for noting accurate attendance5. Social media platforms are tapping into this technology as well, allowing you to automatically tag people in your posts6. Of course, there are still plenty more applications to mention, but these are the main ones where AI facial recognition has made itself known and useful.
AI facial recognition is a paradox. Like any other AI innovation, this was created with a valiant purpose. As mentioned before, it has been helping various fields in various ways. However, not everyone who can use this system shares the same good intentions. AI face recognition helps fortify your security yet it can also compromise your identity.12
You may be familiar with deepfakes, those malicious images and videos circling around the Internet that were made from "stolen" faces. Without AI facial recognition, they wouldn't have been born since it's the essential feature that enables deepfake technology to extract a person's face and replace it with another's. Aside from that, hackers can fool AI face detectors into believing that a deepfake photo or video is authentic, so the system may grant access to the impersonator, endangering your data and assets.2 There are also concerns about large-scale database breaches and leaks. You may have unknowingly fed an unreliable AI facial recognition app with your media files and it may sell or spread your information to shady recipients. Some databases even add your face without consent. However, even if a database is indeed trustworthy, there's still a risk of hackers breaking past its defenses to steal your data.12
Who would have thought that technology would advance so much that it could recognize people's faces? Since its invention in the 1960s, face recognition has become a great addition to many industries and it's now even better with the recent introduction of AI into its system1. It offers stronger security, more accurate identification, and higher efficiency in most of its applications2.
Nonetheless, AI facial recognition is not as bulletproof as you think it is. Even with this protection in place, it's not impregnable and hackers can still break its sturdy walls3. Governments and major corporations must guard against this possibility with properly enforced laws and more security measures4. You as a regular user should also keep your guard up and only share your visual data, and any data for that matter, with credible sites and platforms5.