Facial recognition software makes our lives easier but they have also opened the debate about where to set the limits to data privacy. Credit: NEC.
ISABEL RUBIO | Tungsteno
Many users of a latest-generation smartphone can now unlock their phones with their face. They can also withdraw cash at some ATMs or use their faces at airports to access the boarding area and get on an airplane. They can even access concerts and other events without having to carry a ticket. Facial recognition systems are part of our daily lives and are presented as a tool capable of making our lives more comfortable. But they also give rise to the following debate: To what extent do these techniques interfere with our right to privacy or intimacy?
There are already countries that use this technology to monitor their people. In China, police use facial recognition glasses to identify suspects, and nearly 170 million video surveillance cameras are installed throughout the country. Efforts are also being made to combat truancy with facial recognition systems at the entrances of some schools. And it is even being used in the public restrooms of the Temple of Heaven in Beijing to ensure that toilet paper is not wasted. A machine dispenses a 60-centimetre piece of toilet paper to a user, scans their face and does not allow them to use any more until nine minutes later.
Facial recognition is used to access the airport, but it also allows people to be identified in any context without them noticing. Credit: Shanghai Pudong Airport.
Limits on the use of technology
In contrast, there are places that try to limit the use of facial surveillance by public authorities. San Francisco became the first city in the United States to ban facial recognition technology last May. Other places are also considering preventing the government from using this type of system, such as Oakland and Berkeley, California, and Somerville, Massachusetts.
And there are even individuals who have taken action. A man has taken the Welsh Police in the United Kingdom to court. "I popped out of the office to do a bit of Christmas shopping and on the main pedestrian shopping street in Cardiff there was a police van. By the time I was close enough to see the words 'automatic facial recognition' on it, I had already had my data captured by it. That struck me as quite a fundamental invasion of my privacy," he told the BBC.
This technology is so controversial precisely because it is able to identify people in any context without them noticing. For example, in 2018, U.S. singer Taylor Swift used facial recognition software at a concert to detect harassers among her audience without warning attendees. In addition, privacy experts warn that there is a large amount of information that can be extracted from the face or by combining different facial recognition systems. For example, what clothes someone usually wears, if they wear makeup, if they have any visible disease on their face or even if they are Catholic because they goes to Mass every Sunday.
Biometric identification techniques are also being used for the investigation of rare diseases. Credit. Credit: Paul Kruszka, et al.
Gender and race biases
Added to this is the fact that this type of system usually applies a series of biases. For example, the facial recognition system used by the Welsh police in the 2017 Champions League Final failed to identify nine out of ten people, according to technology magazine Wired.
These errors mainly affect women and black people because the algorithms are trained with massive data sets that typically include a much smaller percentage of images of women and dark-skinned people. A 2018 test by the American Civil Liberties Union (ACLU) using Amazon's facial recognition system shows the presence of these biases. The system falsely matched 28 out of 535 Members of Congress with police suspects. Most false matches corresponded to photographs of people of colour. They wrote at that time warning about the danger of the use of this system by the authorities, because "an identification — whether accurate or not — can cost people their freedom or even their lives."
A useful tool for investigation
But these systems can also be particularly useful —for example, in the search for people with unknown whereabouts. Last year, the New Delhi police used facial recognition to find more than 3,000 missing children. In the field of health, the National Human Genome Research Institute is using these techniques to detect some rare genetic diseases.
With its pros and cons, facial recognition technology has become a key challenge for society and legislators in the coming years, as indicated in a report by the influential research centre AI Now Institute. In fact, the market for facial recognition technology continues to grow, confirming this trend. In 2017 it reached $1.4 billion, according to Statista. This year the figure is expected to rise to $1.9 billion and continue to grow to $3.1 billion by 2022.
· — —
Tungsteno is a journalism laboratory to scan the essence of innovation. Devised by Materia Publicaciones Científicas for Sacyr’s blog.