Facebook (Meta) announced with great fanfare a month ago that in December of this year it will stop using facial recognition in the tagging of photos on its social network and that it will erase the records of 1 billion users. However, company sources acknowledged to The New York Times that DeepFace, the algorithm that had been trained with the images of all those people will not be removed. And that Meta reserves the right to develop new applications in the future based on the processing of biometric data. That is, it is removed from facial recognition, but not actually removed from facial recognition.
Why ? Because the development of the metaverse is based, to a large extent, on that technology. The Cambria project, the virtual reality glasses that will allow us to immerse ourselves in the new world that Mark Zuckerberg wants to superimpose on the physical one, need to collect biometric data from users to reproduce them in the digital environment. The idea is to capture body and facial movements so that the avatars are as realistic as possible. The fear of many analysts is that this is a new way for companies to collect and exploit a new range of user data that until now was forbidden to them.
If facial recognition is going to continue to play an important role in Meta, why then announce that it is being abandoned? The statement was released less than a week after Zuckerberg revealed that Facebook was being renamed Meta, which in turn occurred as internal document leaks from his former employee Frances Haugen dominated accounts of the company. The change of name and the entry into the scene of the metaverse, on the one hand, and the abandonment of a technology as controversial as facial recognition, on the other, could turn the tables.
I'm very proud of the work – both internally, and from the many privacy advocates working on these issues – that has led to today's announcement that Facebook will be shutting down its facial recognition system and deleting over a billion face prints. https://t.co/ZifHIZd1U6
— Kevin Bankston (@KevinBankston) November 2, 2021
The news was in fact applauded by some analysts and organizations. Not surprisingly, companies such as IBM, Amazon or Microsoft had made similar decisions in recent times, although in their case they were prompted by the police application of these systems. The reasons given by Facebook had to do with social responsibility. “There is a lot of concern about the place social recognition technology should occupy in society and regulators are still in the process of passing concrete laws on its use,” Meta Vice President of Artificial Intelligence Jerome Pesenti said in a statement. “We believe that limiting the use of facial recognition to a limited number of cases is appropriate.”
Asked by Newsfresh about Additional reasons for canceling the use of this technology in the identification of photos and on the role it will play in future products and services, the company confines itself to the official Pesenti statement.
Because right now?
The message did not take long to provoke mixed reactions among experts. Kate Crawford, a well-known researcher of the social and political implications of artificial intelligence, called the fact that the company did not close the door to continue using facial recognition as “very goal.” “They imply that they are withdrawing it because they are very concerned about ethics but at the same time they are announcing that they are going to build a virtual reality that is based on facial recognition and many other surveillance sensors. It is totally contradictory ”, says the philosopher and privacy expert Carissa Véliz.
So Facebook is deleting one billion facial recognition scans, but it's keeping DeepFace, the model that is trained on all those faces. Note that “the company has also not ruled out incorporating facial recognition into future products.” Very meta. 👀 https://t.co/8ntPg5Hyf9
— Dr. Kate Crawford (@katecrawford) November 2, 2021
Facebook has already received calls of attention regarding the use of this technology, which identifies and tags users in the photos and videos posted on the social network. Although that function is, since 2019, optional. That year, Facebook had to face the largest fine ever imposed: $ 5 billion demanded by the US regulator (FTC) in the wake of the Cambridge Analytica data breach scandal. Among the arguments put forward by the regulator, Véliz recalls, were concerns about the erosion of privacy related to the application of facial recognition in its services.
In February of this year, Facebook paid $ 650 million to settle a lawsuit filed against the company in Illinois, the only US state in which it is illegal to collect biometric data without the explicit consent of the people. “How strange that after those two setbacks they did not cancel facial recognition, especially if they say they do so because they are concerned about its impact on society, and they do it when Haugen airs dirty laundry of the company,” says Véliz.
“There are many reasons why we should be skeptical about Facebook's announcement that it will withdraw the use of facial recognition,” he says for his part Ella Jakubowska, coordinator of the facial biometrics program at the Brussels-based NGO European Digital Rights Initiative (EDRI), a Brussels-based NGO working to defend human rights in the digital age. “While it is positive that they erase their records of many people's facial patterns, presumably they have already extracted the value they wanted from those images, in addition to having used our faces to train their systems for profit.”
The metaverse needs to read us
As detailed by Evan Selinger, philosopher of technology and expert in the ethical derivatives of the use of facial recognition, the gaze detector that the glasses will incorporate will be essential to bring realism to the metaverse, allowing for example that users can establish direct eye contact (which cannot be done in video calls: or you look at the camera or the screen). “Gaze tracking is the Holy Grail of advertising. Going beyond commercial internet clicks could make a fortune one day, ”he wrote in The Boston Globe .
From a technical point of view, offering a totally sharp digital image in virtual reality glasses can make the user dizzy . We are used to seeing well where we aim our eyes since the rest of the field of vision becomes more blurred as we move away from the center. To detect where that point is, you need a gaze detector. The problem is that it does not only work for that. “By following your gaze, many things can be inferred: from what catches your attention to how fast you are thinking, since the time it takes to react to something indicates things about your cognitive function. It also reveals what you like and what you don't like or what attracts you, because the pupils dilate in those cases ”, explains Véliz. Of course, iris analysis, which describes a unique pattern in each human being, reliably identifies us, just like our fingerprint does.
What credit can then be given to the announcement of the withdrawal of facial recognition? “Facebook's business model is based on knowing you intimately, and that hasn't changed,” argues Jakubowska. We should not trust that a company that just a few months ago launched glasses with integrated cameras will create a metaverse in which privacy rights are respected. ”
You can follow Newsfresh TECHNOLOGY at Facebook and Twitter or sign up here to receive our weekly newsletter .
- Do you remember the faces but not the names? A nap could help you, according to science
- The 'blackberry' dies: signed the death certificate of the mobile that wanted to reign
- A solar eclipse at the north pole excited auroras in the southern hemisphere
- League of Legends, Netflix and Disney + promote the Canary Islands in the animation industry