Updated: Feb 12
Breaking down the social implications of facial recognition technology
In May 1999, Wang Hua gave birth to a boy in the Chinese province of Sichuan in Dujiangyan. At the age of five, the child was kidnapped while playing near a construction site where his father worked. He was sold to child traffickers for the equivalent of 2000 dollars.
After 20 years, Chinese detectives were able to use facial recognition technology to match the picture of the toddler given by his mother to a 20-year-old student living in Guangdong province. A DNA test then confirmed to Wang that it was her baby, now a grown adult.
According to the National Crime Records Bureau, one child disappears every eight minutes in India and China due to the high volume of child trafficking. Facial Recognition Technology (FRT) has proved to be extremely beneficial to police forces to find lost children in such populated parts of the world where in recent news, nearly 3000 children were found in just four days.
Surveillance has become a major part of society today. Police are always looking for technologies to enhance their security methods. Facial recognition technology can identify the facial image of a person and compare it with other pre-existing files of the person. These files can be found in a variety of places such as government records, mugshots, or social media. This technology has become a game-changer in the police investigation as it helps identify lost children, fugitives, terrorists, and frauds. For instance, police in New York identified over 10,000 people with more than one driver’s license with the use of FRT since 2011. Such identification thefts are becoming one of the fastest-growing crimes in the world. Emerging technologies like facial recognition have proven to aid surveillance systems for civil forces around the world. It is rapidly gaining popularity in the market due to its ability to provide fast results as compared to other AI technologies. FRT grants a quick and relatively secure system that can verify a person in just a few seconds from a large distance.
While we cannot deny the advantages of this emerging technology, it is crucial to discuss its consequences and be ready with solutions.
AI-based technologies are quickly becoming accessible to almost all governments in the world that have the ability to recognise our faces just as good as that of a 5-year-old. In the UK, however, the use of this technology has been criticised severely by its citizens. Events such as London’s annual Notting Hill Carnival and major football games were seen deployed with specially equipped vans using FRT. Those attendees who refused to show their faces at such events were penalised with a fee. In China, this technology is openly used to name and shame citizens. Faces of citizens are displayed live on big screens viewed publicly even for offenses like jaywalking. Chinese police are using sunglasses equipped with facial and gesture recognition technology for real-time analysis of faces. Hong Kong’s citizens continue to protest for the right to be shapeless and anonymous among the crowd with the protest's striking phrase of what roughly translates to Be water.
FRT raises the ultimate question, just like with many other frontier technologies such as AI- where is the balance? Safeguarding rights and freedom of people has never been more important.
Social networking sites such as Facebook secretly selling user information to advertising companies exposed a similar problem with privacy issues. It is perturbing that some companies can commercialize personal information of consumers without consent. Similarly, FRT is also no longer only about security-checks by the government but has also entered into commercial businesses, such as doing online payments or check-in at airports using one’s face. Amongst China’s many artificial intelligence startups, one of them is by a company called Yitu, which is creating an intelligent city where facial recognition is part of daily routine, for instance, opening a door with the face of the owner. This could even extend to revealing intimate details about an individual’s personal life over an extended timeframe.
Recently, companies such as Diebold have developed ATM where you can transact money with face of the user. However, the user must store a copy of his or her biometric measurements to use Diebold’s ATM which is highly controversial. Although the use of such commercial services is voluntary, users of such services are not made fully aware of the risks. For a long time, taglines of major brands using FRT commercially have stated that it is “hundred percent safe and secure,” while in reality, it is not yet fool proof against situations of misuse and exploitation. This raises an issue of a lack of trust and reliability with such technologies. Facial recognition is an extremely promising technology; however, we do not want it to lose public trust or interfere with fundamental rights.
FRT must be commercialized carefully with risk communication becoming an active part of research and practice. Hiding information from the public will not only be deceiving but also receive more criticism in forms of protests such as observed in China and the UK. Users should be able to easily find information on websites or info centres regarding the risks involved in case of misuse or hacking. People should also be able to know the record of their faces in data saved by companies making use of this technology. Last but not the least, one must have the option to choose alternative methods such as fingerprints that may or may not be more time consuming if he or she is hesitant towards facial recognition, for instance, in cases of identical twin siblings. Transparency is the key element that can go a long way to establish brand loyalty. Government and commercial companies need to be transparent with the way they use customer data. Governments are also encouraged to make use of information boards at airports and other locations where facial recognition is used for surveillance. Such info boards must contain information about the fundamental rights of a person in such circumstances or alternative methods available for checking-in at the airports etc. Nonetheless, it is essential to strike a balance between presenting about the pros and cons of the technology. In order to test such communication schemes, companies like Yitu can be used as a pilot program in China. These schemes can be incorporated in the marketing of potential intelligent city projects and evaluate customer feedback.
Like most AI technologies, FRT is dependent on the ability of a machine to learn. The machine must be fed with plenty of data such as facial angles, haircuts, skin colours, etc. to come up with more accurate results. However, so far, the facial recognition technology is better calibrated to white men and thus, results in producing a higher number of incorrect results for women and people of colour as compared to the white population. This misidentification of faces increases the chances to put certain groups to be associated with crimes significantly. Although such inequalities are slowly improving. According to the National Institute of Standards and Technology, the modern algorithms miss a correct match only 0.2 percent of the time and got 25 times better at identifying a person between 2010 and 2018.
Facial recognition technology is one of the most promising AI technologies that have the potential to transcend the digital era, though, in unprecedented ways. It is a dire need for law enforcement officials to develop rules and regulations in compliance with the obligation for personal data as well as the right to privacy.
We no longer want emerging technologies to time and again fraught with issues such as civil liberties and its ethical sustainability over the long run.