Is Facial Recognition Racist?
Facial recognition technology has quickly become a staple in biometric security and surveillance. The technology can be found everywhere from smartphone security to restaurant ordering systems to recall user preferences. But, researchers have long been critical of one weakness the current technology has, dark faces.
For years, even the best facial recognition technologies have had a bad record telling dark faces apart. Incidents at airports are only the beginning, multiple reports here, here, and here based on both independent and government research all agree that women and people of color are misidentified up to 10 times more than white men. A highly respected study done in 2018 by Joy Buolamwini and Timnit Gebru shows how current technology may be fundamentally flawed when it comes to gender and racial disparities.
Anecdotally, the most common use for facial recognition is in smartphone security. Apple got rid of its fingerprint scanner years ago and users must rely on their facial recognition. But when the iPhone X was released there was a slew of Chinese users that claimed the security software cannot tell one Chinese user from another. Apple’s response to the numerous instances was that it was not a bug but a fundamental error to the technology. No solution other than a refund was offered for the affected users.
But perhaps the most telling sign of how fundamentally flawed the technology is when Google’s facial recognition labeled two Black people as gorillas in 2015. The incident was solved by removing gorillas as an identifiable group. Even as recent as 2018 they never solved the underlying issue. Google lens continues to not be able to identify gorillas as well as other monkeys implying Google engineers never found a way for their technology to not label Black people as apes.
It is important to remember that technology is not inherently racist, but we must be mindful of what technology is used for and the context of the concerns being raised. There are widespread fears that facial recognition will be used for mass surveillance and may be used to identify people exercising their rights like in the case of Black Lives Matter protests.
Company Responses
Black Lives Matter movement leaders also brought attention to the use of technology by law enforcement that is inherently more detrimental to minorities. IBM has announced that it will no longer develop facial recognition technology. Amazon has also come out and banned police use of its facial recognition technology for one year.
In a letter IBM sent to Congress, the CEO stated they will not allow the misuse of their technology to violate human rights and that IBM strictly opposes the use of facial recognition “for mass surveillance, racial profiling, violations of basic human rights and freedoms.”
The CEO also suggests working together with the government for a more human approach to police reform, more transparent use of technology, and expanding opportunities for communities of color through education and training.
IBM advocates for the responsible use of technology through transparency and open communication while also holding those that use technology accountable for any misuse.
“We believe now is the time to begin a national dialogue on whether and how facial recognition technology should be employed by domestic law enforcement agencies. Artificial Intelligence is a powerful tool that can help law enforcement keep citizens safe. But vendors and users of Al systems have a shared responsibility to ensure that Al is tested for bias, particularly when used in law enforcement, and that such bias testing is audited and reported,” IBM stated.
Amazon will not be going as far as IBM and discontinuing all development of facial recognition, Rekognition, but it will prohibit police use for a year. The short blog post that Amazon released did not specifically mention possible human rights violations or racial discrimination, supposedly because the services will continue to run for other organizations. But it is likely Amazon decided to suspend police partnerships because of pressure from Black Lives Matter leaders, as well as IBM’s decision, made just two days prior to Amazon’s.
Amazon did write about the need for more stringent regulations for the ethical use of emerging technologies like facial recognition. “We’ve advocated that governments should put in place stronger regulations to govern the ethical use of facial recognition technology, and in recent days, Congress appears ready to take on this challenge. We hope this one-year moratorium might give Congress enough time to implement appropriate rules, and we stand ready to help if requested,” Amazon wrote.
Future of Facial Recognition
Although IBM has pulled back from facial recognition entirely, the biggest players, Google, Apple, Facebook, Amazon, and Microsoft continue to develop and sell the technology.
Facial recognition companies and advocates believe that with time the technology can accurately identify faces regardless of what gender or race they are but nobody seems to have any idea when that time will come. As far as researchers can determine, the majority of facial recognition advancement occurred in the speed of identification, not in accuracy.
We may have to wait until someone creates an entirely new method of facial recognition for truly accurate identifications.
As public opinion shifts and companies take matters into their own hands, lawmakers may have to create laws regulating the emerging technology.
It is unlikely that the technology will be banned completely even from law enforcement as there are highly useful and legitimate uses for facial recognition technologies. It is frequently used to solve human trafficking and child abduction cases. Many petty crimes such as theft are also difficult to solve and are largely accepted as lost cases, but facial recognition has helped many police departments solve those cases as well.
What is needed with this and any emerging technology is proper and informed regulation. Whether or not lawmakers will work with industry leaders to craft effective regulations that do not encroach on innovation is still to be seen.
