Social media as the end of privacy
Facial recognition is made possible in large part by the willingness of average people to cast their digital likenesses out into the world wide web, such that anyone can potentially be looked up via a search engine. Maybe you have a LinkedIn or Facebook profile that you use your real name on. Well, that’s one way someone can find you with Google. But now, technology’s ability to intelligently scan a face has improved greatly. People have unknowingly provided all the fodder the technology needs – millions of images of their faces at the most flattering angles – creating a database that allows the tech to learn and improve. Plus, the more angles there are of a given face, the easier it can be identified. Like deepfake videos – where a face can be convincingly slapped on a new body digitally, or be made to “say” things that were never said – the rise of face ID is just one more way in which your face is becoming about as charted as a map of Manhattan.
And then there’s something like Deep Face, Facebook’s own facial recognition tech, which has been around since 2015.The technology is mostly so users can tag unidentified people in their photos – with apparently 97.35% accuracy.
But while Facebook has no problem face ID-ing people from images that are uploaded to its own site, it apparently has a problem with other companies doing it.
When companies freely make use of your images
The startup company Clearview, which is currently only available to law enforcement, is one of the most discussed current practitioners of facial recognition. They already claim to have scraped over three billion images from sites like Facebook and YouTube, and millions more websites as well. The legality of that practice is hotly disputed, but meanwhile, hundreds of law enforcement agencies, FBI included, are making gleeful use of Clearview’s technology to identify and track down suspects. Twitter, Facebook, and LinkedIn have fought back, asking Clearview to cease and desist. Clearview has invoked the first amendment, arguing that what they do isn’t really all that different from the web searching already done by Google, and the battle wages on.
Clearview’s own philosophy is murky at best, but they seem committed to aiding law enforcement, even when it exists in authoritarian countries with poor human rights records.
Accuracy of face-ID tech is also dubious. While Clearview publicly proclaims an astonishing and debated 100% accuracy, Amazon’s own superbly named Rekognition face-ID software, which is also used by law enforcement, showed in a test to misidentify 28 members of congress as people who’d been arrested for a crime. Moreover, false matches seem to occur more frequently with people of color. But until legislation prevails, the technology and its uncertain accuracy will continue its uneasy work.
Tech regulation has proven to be notoriously difficult so far, and any law of the land on facial recognition will be no exception. In the United States, facial recognition tech is being fought at the local level. Take San Francisco, which banned facial recognition – meaning police and government agency surveillance – in May 2019, but then ran into trouble over what to do with Apple’s Face ID system, which is used instead of a passcode to unlock personal devices and is present in many of their latest phones.
The European Union, which had recently been considering a five year ban on the technology in order to assess its impact, has more recently elected to let individual countries decide their own rules on this very divisive issue.
Universities have also taken individual stances on the technology. Some are already using it because they believe it will increase safety around campus and in dormitories. Some have declared they won’t use it, and then there are others who aren’t using it yet but are interested in it.
Behind all these face-ID anxieties are pretty obvious dystopian connotations. Indeed, privacy as we know it is at risk. The paranoia around face ID is already having what the ACLU refers to as a “chilling effect” on daily public life. There’s the rub: despite its do-gooding quotient, face ID is a step toward a suffocating authoritarianism.
Facial recognition in China
If you want to see what the future of face technology might look like, take a peek at China. Facial recognition and other forms of surveillance have already become ingrained into Chinese society to the point where people may hardly think about it. It is casually present in some shops, hotels, and transportation stations, where customers knowingly interact with face-ID machines. In December 2019, a measure was passed to make new mobile phone users submit to face-ID scans – ostensibly to prevent phone theft. While such precautions do have supporters, the rule has also upset many Chinese, who fear that there aren’t many vestiges of privacy left.
Beyond face value
Like most tech, facial recognition has some upsides: in this case, convenience and crime-fighting. But unlike most other tech, if facial recognition takes over, we won’t have much choice about how it’s used or not. Whereas with apps, one still clings to that shred of online privacy protection known as “opt in/out”, there is no reasonable opt in/out if face ID becomes part of everyday life. And while face ID should be more secure than a regular password, if it is hacked, you’re in trouble: you can change a password, but it’s not as easy to change your face.