The ability to digitally identify faces has always been an important step in the development of artificial intelligence. In recent years, enormous strides have been made in the advancement of facial recognition technology, and along the way important questions have been raised about its future role in our everyday lives.
Many organizations such as the American Civil Liberties Union have expressed concern that privacy could be affected if restrictions are not established. They fear this technology could be exploited by government agencies and private corporations to manipulate consumer trends or to keep constant surveillance on individuals.
Sure, a “total surveillance society” sounds like something taken out of dystopian science fiction, but technology has evolved so rapidly in the past decade that it is practically impossible to stay away from facial recognition software nowadays.
From Facebook to security systems in airports, and even cameras in shopping malls, your face is constantly being “read.”
There are basically two types of facial recognition software. The first one just analyzes demographics: It sees a face and determines age, race, and gender. It’s used mainly as a commercial identification tool, as it helps companies understand who is interested in their products or services.
The second type is for identification purposes, so it doesn’t determine what you are but who you are. In other words, it can connect a stranger’s face to a name. And this is the type of facial recognition people worry about.
Privacy advocates argue that this kind of accuracy—combined with social media provided data such as location and IP address—could end up compromising our individual privacy.
Their concern, however, is not entirely directed at government agencies, even though different agencies have been using facial recognition technology for decades. Their concern is mainly directed at tech giants like Facebook or Google, who currently lead the way when it comes to facial identification software.
Government agencies such as the FBI or Homeland Security rely mostly on ID photos and mugshots for facial identification, while social networks like Facebook rely on their enormous database of names and photos. For example, every time one of Facebook’s 1.8 billion users uploads a photo and tags a person, they are helping the facial recognition algorithm by showing how someone looks in different angles and lighting.
In fact, all this information provided by its own users, along with its highly-advanced software, has made Facebook reach near-human performance. By “looking” at a photo, its software can accurately identify a person 97.25% of the time—humans will do it 97.5% of the time.
Facebook has repeatedly said it does not collect all user data and the data collected is only used for benign purposes. Still, many organizations consider the social network a threat. They argue that government, through its law enforcement agencies, can issue warrants for any information available on Facebook. And even though the FBI has claimed they do not store Facebook photographs in their facial recognition database, there is nothing in the law to prevent them from doing that.
Advanced facial recognition systems are fairly new, but many companies are working on ways to improve this technology and incorporate it in our everyday lives. So, next time you upload a photograph of yourself or tag someone, remember you’re most likely not sharing it with only your friends.
By Alonso Montano