Recently in the news was the attention-grabbing headline: “Attorney says facial recognition got her kicked out of a Rockettes show.”. The lawyer in question happened to work for a firm that was suing the owner of the venue (MSG), though the lawyer did not work on the case. Regardless of how it happened, this opens up a very interesting dialogue as to ethical and legal questions regarding the use of facial recognition for more than simple passive surveillance. Is it ethical? Is it legal? Should be it? No matter what side of the political aisle you’re on, there is probably a clearcut answer – except no one is going to agree.
There are two opposed but clearcut answers on this situation, for reasons that are almost identical. First, that the use of facial recognition technology to ban someone from a private business should be allowed. After all private businesses have the right to refuse service to anyone. So if that is done via a person remembering a face, a printed picture on a wall, or a computer that goes “Beep” – it’s the businesses prerogative. Conversely, facial recognition technology should be banned in private businesses because businesses are not allowed to discriminate. Additionally, the use of such technology is an invasion of privacy.
The private businesses rights to refuse are simplistic in concept but complicated in execution. “The Right to Refuse Service and Its Implications on Society” by Duke’s Undergrad Law Magazine provides a solid summary of the topic and some of the recent court cases. The waters get very muddied when talking about the output product of a business (ex: Should a cake baker be required to make a cake for a gay couple, if that baker believe that to be immoral?). However just focusing on the current situation of “admittance” to a business, the Civil Rights Act of 1964 is fairly simplistic: You can’t kick someone out for race, skin color, religion, sex or national origin. It doesn’t say anything about preventing a business for giving a lawyer the boot.
At the most basic level, this shouldn’t be a big issue. However, the use of facial recognition technology is ripe for abuse on so many different levels that we should be worried now. In fact, 2 years ago Last Week Tonight with John Oliver covered this exact topic.
As businesses have liberties, so do individuals, and the broad usage of facial recognition is a huge impingement into everyone’s right to privacy. Not only will your data be collected, without your consent, it will likely be stored and catalogued in a giant database with millions or billions of other records. Minority Report (yes, the Tom Cruise movie) showed a “lovely” example of the use of similar technology in advertising. Imagine if there was a database of every store you’ve walked into with exact dates, times and purchase details? Most of that already exists with club/savings cards but those are at least opt-in. With facial recognition there is no way to even know if it’s being used, let alone opt-out. And don’t worry, it’s 100% certain that data will be sold.
Getting back to our original concern of businesses banning customers there are three major technological concerns to keep in mind. First: The technology is not that accurate. It’s not recent news that “U.S. government study finds racial bias in facial recognition tools” and “Research: Students of color at greatest risk for facial recognition errors.” If you are not-white, then it is likely the facial recognition technology will be error prone at a greatly increased rate… and remember what we said just a few paragraphs ago about the Civil Rights Act of 1964? Can’t kick someone out for skin color. Suddenly it’s not so simple. Misidentification is a real problem and could lead to people being kicked
The second major concern is the use of algorithms. Today the use is a adding specific faces to a database for banning. But what if it wasn’t just a specific face, what if they use algorithms to look for potential signs of shoplifting? Or maybe your local American multinational luxury jewelry and specialty retailer just doesn’t think your income is suitably high enough for their store? Those algorithms are made by people, and people have biases. The algorithms are also so complicated that people don’t understand how they fully work, already today that’s the case and it has many issues: “We don’t understand how YouTube’s algorithm works—and that’s a problem”
The final major technology concern is what the backend database is, who’s been added and why? You can kick someone out of your store for any reason… so long as it’s not discrimination. So it’s not the act that matters so much as “why” someone acts. In legal terms, the Mens rea or “the person’s intention to commit a crime.” How do you prevent a business from just banning anyone at random? Today it is hard, but with a facial recognition database you can easily ban anyone anytime. Maybe a guest stole something from the shop – good reason to ban them. Maybe you just don’t like the way they look. Or if you want to take it a step further, what if it wasn’t just a one store database? Or even a one business database? What if being banned (accidentally, misidentification, or purposefully) from a single Target also got you banned nationwide from every target, K-Mart, Sam’s Club, Costco, Walmart and Safeway – because they all use the same database?
Technology always moves faster than the law. From a business rights perspective, it would be nice to assume that everything will be done with the technology. However, there are already rumors that MSG, the same venue owner, has been using the same technology to target and harass others. Outside of business, the Police have a poor record regarding the use of this technology, going as far as ending up with the wrong people getting arrested. While the use of facial recognition is not easy to define cleanly, it’s clear it’s already being abused.