Facial Recognition and the Danger of Automated Authoritarianism

Hundreds of police departments are using facial recognition technology without oversight.

Clearview AI, a tech startup, has created an app that enables law enforcement agencies to match photographs to its database of over 3 billion photos scraped from millions of public websites including Facebook, YouTube, Twitter, Instagram, and Venmo. For comparison, the FBI's photo database contains only 640 million images. According to The New York Times, some 600 law-enforcement departments, including federal, state, and local agencies, have already used Clearview AI's technology. "You take a picture of a person, upload it and get to see public photos of that person, along with links to where those photos appeared," explained the Times.

Moreover, Clearview AI's technology doesn't need straight-ahead mugshots to work effectively. "With Clearview, you can use photos that aren't perfect," Detective Sgt. Nick Ferrara of the Gainesville, Florida, police department told the Times. "A person can be wearing a hat or glasses, or it can be a profile shot or partial view of their face." No outside tests for the app's accuracy have been publicly reported, but the company claims in a 2019 FAQ to users that it:

…has the most accurate facial identification software in the world, with a 98.6% accuracy rate. This does not mean that you will get matches for 98.6% of your searches, but you will almost never get a false positive. You will either get a correct match or no results. We have a 30-60% hit rate, but we are adding hundreds of millions of new faces every month and expect to get to 80% by the end of 2019.

The current Clearview AI app works basically as an investigative tool helping police identify perpetrators or victims after a crime has occurred. The company is, however, developing facial recognition software that would make it possible for wearers of augmented-reality glasses to ID folks walking down a street in real-time. Of course, such a technology could easily be harnessed to networked surveillance cameras so that government agents could track where a citizen is and with whom that citizen is interacting.

"Facial recognition is the perfect tool for oppression," write Woodrow Hartzog, a professor of law and computer science at Northeastern University, and Evan Selinger, a philosopher at the Rochester Institute of Technology. It is, they persuasively argue in Medium, "the most uniquely dangerous surveillance mechanism ever invented." Real-time deployment of facial recognition technologies would essentially turn our faces into ID cards on permanent display to the police.

"I've come to the conclusion that because information constantly increases, there's never going to be privacy," Clearview AI investor David Scalzo told The New York Times. "Laws have to determine what's legal, but you can't ban technology. Sure, that might lead to a dystopian future or something, but you can't ban it."

In fact, several cities have already banned police use of facial recognition technologies. And members of Congress are also now waking up to how the widespread use of this technology could impair our civil liberties. "In November [2019], Sens. Mike Lee (R–Utah) and Chris Coons (D–Del.) introduced the Facial Recognition Technology Warrant Act," reports Reason's Scott Shackford. "The bill would require federal officials to seek a warrant in order to use facial recognition technology to track a specific person's public movements for more than 72 hours."

Permitting police to track a person using facial recognition for less than three days without a warrant seems constitutionally questionable to me. After all, in Carpenter v. United States (2018), the Supreme Court required that the police obtain a warrant in order to search a person's cell phone location data. "A cell phone—almost a 'feature of human anatomy'—tracks nearly exactly the movements of its owner," Chief Justice John Roberts wrote in his majority opinion. "When the Government tracks the location of a cell phone it achieves near perfect surveillance, as if it had attached an ankle monitor to the phone's user." Since faces are actual features of human anatomy (no ankle monitors needed), surely a warrant should be required for police to track citizens by their faces.

Harvard cybersecurity expert Bruce Schneier argued in a recent New York Times column that merely banning facial recognition is not enough to protect ourselves from government snooping. Ubiquitous mass surveillance is increasingly the norm. "In countries like China, a surveillance infrastructure is being built by the government for social control," he wrote. "In countries like the United States, it's being built by corporations in order to influence our buying behavior, and is incidentally used by the government."

The data exhaust constantly emitted by our cell phones, license plates, digital payments, and credit cards is an intimate and comprehensive record of our lives. Right now easy government access is hampered by the fact that the myriad private databases tracking us are disparate and unconnected. However, it is not hard to imagine how automated authoritarianism could arise quickly in the wake of a massive terrorist attack as frightened citizens set aside concern for civil liberties and give in to government demands for access to all of the data that has been privately collected on each of us.

Given the growing prevalence of both government and private surveillance, Schneier argued, we Americans "need to have a serious conversation about … how much we as a society want to be spied on by governments and corporations—and what sorts of influence we want them to have over our lives." To forestall a dystopian future, setting strict limits on government use of facial recognition technology is a good place to start.