For decades, Facial Recognition has been a stuff of science fiction but in many countries, it has already become a part of daily life. Though the technology dates back to 1960’s yet there is always something new that catch holds our attention. Facial Recognition has made it possible for our iPhones to identify us, Facebook to tag us in our pictures. Moreover, this technology is widely used for automation, finding missing people, check-in at the airports, login to mobile apps globally. It has the ability to replace passwords and fingerprints thereby increasing the security. It is assumed that Facial Recognition will speed in enforcing law and order. Facial Recognition is indeed a powerful tool. However, with great power comes great responsibility.
Facial Recognition’s dirty little secret: millions of online photos scraped without consent!
Now, a photograph is not just a photo, it is the point of identity of an individual. U.S uses Facial Recognition for reducing cross border crime and undocumented migrants which violates Human Rights. Facebook’s
DeepFace is alleged to be collecting and storing face recognition data without user’s consent which clearly violates Bio-metric Information Privacy Act. The extent of affliction may be unclear in the U.S., but in China overuse of Facial Recognition is the testimony of how this great a technology can get out of hands. The Chinese government not only use it for mass surveillance, but also for giving the citizens a Social Score. This score is based on economic and social reputation and can be affected by bad driving, posting fake news online or even buying too many video games. If you get a good score, you might get the privilege to avail discounts on bookings but if you get a bad score, you may be banned from boarding flights or trains. In a way the government is using it publicly to embarrass citizens.
Facebook’s AI lab can recognize face with 97.25% accuracy which is 0.25% less than the accuracy with which humans recognize a face. It is interesting to know that Amazon’s Facial Recognition technology often mistakes dark-skinned women for men!! Yes, you read it right. Study suggests Facial Recognition to be racist and gender-biased. Well, this is not some sort of anti-racist technology. The real reason behind this discrimination is that they don’t have lot of database of African People.
So, can we really conclude that Artificial Intelligence based Facial Recognition is really worth replacing other static bio-metric methods like fingerprints, iris scan, palm recognition and many others. As Brian Reed has correctly quoted “
Everything is designed. Few things are designed well.”