By banning face recognition, we miss the point

The whole point of modern surveillance is to distinguish between people in order to treat everyone differently. Face recognition technology is only a small part of the total surveillance system

Essay Writer – Bruce Schneier, American cryptographer, writer and information security specialist. Member of the board of directors of the International Association of Cryptological Research and member of the advisory board of the Electronic Privacy Information Center. The essay was published on January 20, 2020 in the author’s blog and in the newspaper. The new york times.

Communities of concerned citizens across the United States are beginning to ban face recognition technology. Last May, they were banned in San Franciscosoon followed by a nearby Auckland, and Somerville and Brooklyn in Massachusetts (ban may extend for the whole state) In December, San Diego suspended the face recognition program in anticipation of the entry into force of the new law. Forty major music festivals promised not use this technology, but activists call for a national ban. Many Democratic presidential candidates support at least a partial ban face recognition.

These efforts are motivated by good intentions, but the ban on face recognition is the wrong answer to the problem of modern surveillance. Attention to one specific identification method distracts from the nature of the observation society that we are building, where widespread mass surveillance is becoming the norm. In countries like China, the government is creating a total surveillance infrastructure to control society. In countries like the United States, it is created by corporations to influence purchasing behavior, and is used by the government along the way.

In all cases, modern mass surveillance has three main components:

  • identification;
  • correlation;
  • discrimination.

We will consider them in turn.

Face recognition is a technology that can be used to identify people without their knowledge or consent. It relies on the prevalence of surveillance cameras, which are becoming more powerful and compact, and machine learning technologies that can compare the footage with images from the database of existing photographs.

But this is just one of many identification methods. People can be identified at a distance by palpitations or gaitusing a laser system. The cameras are so good they can read fingerprints and iris from a distance of several meters. And even without all these technologies, we can always be identified, because our smartphones broadcast unique MAC addresses. We are identified by phone numbers, credit card numbers, car numbers. For example, China for its total surveillance system uses several identification methods.

Once we are identified, data about our personality and actions can be correlated with other data collected at other times. This can be movement data to “track” a person throughout the day. Or data on purchases, page views on the Internet, with whom we communicate by e-mail or in chat rooms. This may be data on our income, ethnicity, lifestyle, profession and interests. There is a whole industry of data brokers who make a living by analyzing and data supplement about who we are – using observation data collected by all kinds of companies that are sold to brokers without our knowledge or consent.

In the United States, there is a huge – and almost completely unregulated – industry of data brokers that sell our personal information. This is how large Internet companies such as Google and Facebook make money. It’s not just about identification. The main thing is that they are able to create deep profiles for everyone, collecting information about us and our interests and maximally supplementing these profiles. That is why many companies buy license plate data state authorities. That is why companies like google buy medical records, and partly why google bought Fitbit along with all its data.

The whole purpose of this process is for companies – and governments – to differentiate between people and treat them differently. People are shown different ads on the Internet and are offered different rates on credit cards. Smart billboards display different advertisements depending on your profile. In the future, they may automatically recognize us at the entrance to the store in the same way as they do now when they enter the website.

It doesn’t matter what technology is used to identify people. The fact that there is currently no comprehensive database of heartbeats or gait does not make data collection technologies less efficient. And in most cases, the relationship of the ID with the real name does not matter. It is important that they can consistently identify us for a long time. We can be completely anonymous in a system that assigns a unique cookie to each user and tracks his actions on the Internet, but this does not interfere with similar processes of correlation and discrimination. Same thing with faces. You can track our movements around the store or shopping center even without reference to a specific name. And this anonymity is fragile: as soon as we bought something with a bank card, suddenly our real names are attached to what was an anonymous tracking profile.

To regulate this system, all three stages of the surveillance process must be taken into account. The ban on face recognition does not matter if video surveillance systems switch to identifying people by the MAC addresses of smartphones. The problem is that we are identified without our knowledge or consent, and society needs rules when this is permissible and when not.

In the same way, we need rules, how our data can be combined with other data, and then buy and sell without our knowledge or consent. The data broker industry is almost completely unregulated; There is only one law – adopted in Vermont in 2018 – that requires data brokers to register and explain in general terms what data they collect. Large Internet surveillance companies, such as Facebook and Google, are collecting more detailed files on us than they were with the special services of any police state of the 20th century. Reasonable laws will help prevent the worst of their abuse.

Finally, we need clearer rules on when and how companies can discriminate. Discrimination based on protected features such as race and gender is already illegal, but these rules are not effective against modern surveillance and control technologies. When people can be identified and correlated with unprecedented speed and scale, we need new rules.

Today, face recognition systems have taken the brunt of criticism, but by banning them, we miss the point. We need to seriously talk about all technologies of identification, correlation and discrimination. We, as a society, must decide whether such espionage is allowed by the government and corporations – and how we want them to influence our lives.

Similar Posts

Leave a Reply