In the past it was a revolutionary idea which was only possible in action movies where the FBI found the bad guy by searching through footages of security cameras with the help of facial recognition software. Nowadays, this once seen as sci-fi technology has become reality and its use has gone far beyond law enforcement. In fact, it is used by many well-known companies and integrated in products which we daily use. Most mobile phones for instance give you the option to name faces on a photo to subsequently sort out other photos that have the same person in it (Gemalto, 2018).
Facial recognition software is modelled after our own brains ability to recognize faces. However, it has been so far developed that it sometimes performs even better than the human brain. On Washington Dulles airport they caught a fraud who tried to enter the country with a fake passport. The passport was fabricated so well that a person would probable not notice that it is a fake but the facial recognition software did (Newman, 2018).
A human face has numerous distinguishing characteristics by which it can be recognized such as the form of your eyes, the length between one’s nose and mouth or the shape of a hairline (Symantec Corporation, n.d.). The software uses 2D or 3D sensors to capture a face. All these characteristics are than processed by a complicated algorithm to turn it in to digital data. This digital data could then be used to recognize this individual in other pictures or videos (Gemalto, 2018). You could wonder how long this data would stay valid and usable as people tend to change considerably over time. Nonetheless certain aspects of a person’s face stay the same which has makes it possible to still recognize a face even when it has aged. Due to this, the FBI has been able to identify a criminal based on a 20-year-old picture (Symantec Corporation, n.d.).
This also shows the dangerous side of the software. Facial recognition makes it nearly impossible to hide. Even if years pass you are still identifiable by the software and with the high amount of security cameras everywhere you can be tracked almost everywhere. This is of course advantageous when you want to capture a criminal, but it is also a serious violation on one’s privacy. Thereby, it can also be used by stalkers or other criminals. Furthermore, facial recognition software is progressively being used as a security measure for businesses as well as consumer applications. This brings two potential dangers. Firstly, security measures using facial ID can be tricked and are thus not fully reliable. Secondly, data about your face can potentially be hacked or sold to third parties (Williams, 2017). All in all, the development of facial recognition software is very useful in various domains but also comes with serious dangers.
Gemalto. (2018, August 31). The top 7 trends for facial recognition in 2018. Retrieved from https://www.gemalto.com/govt/biometrics/facial-recognition
Newman, D. (2018, September 18). Facial Recognition Software: The Future Is Here. Retrieved from https://www.forbes.com/sites/danielnewman/2018/09/18/facial-recognition-software-the-future-is-here/#4fd94b3e299d
Symantec Corporation. (n.d.). How Facial Recognition Software Works. Retrieved from https://us.norton.com/internetsecurity-iot-how-facial-recognition-software-works.html
Williams, H. (2017, December 18). What are the dangers of facial recognition technology? Retrieved from https://www.techworld.com/security/facial-recognition-technology-what-are-dangers-security-preventions-3669268/
Although hacking the facial recognition software comes with serious dangers, tricking the system has proven te be very difficult. One of the articles you used, mentions that some very skilled hackers tried to trick the iPhone X facial recognition software, but did not succeed. Furthermore, in cases where security is at stake, crossing borders for example, the facial recognition could be complemented with other ways of identifying someone. Facial recognition itself is very hard to trick, but combined with other measures it could potentially be impregnable.