The Worrisome Future of Policing Technology

The Supreme Court decided on Friday that before the government can get your cellphone location records — well, at least more than six days’ worth of them — it must have probable cause to think you did something wrong, and a warrant.

That’s important, because your cellphone is a mini-GPS device broadcasting your location constantly. It provides, as the court majority recognized, “a comprehensive chronicle of the user’s past movements.” Yet, before this decision, law enforcement agencies were gathering this data on countless people without even showing that the person whose information they collected had done anything wrong.

But the 5-4 decision in the case, Carpenter v. the United States, should give you little comfort. The growing use of technology by law enforcement agencies to monitor or target people — particularly people and communities of color — is expanding at head-spinning speed, and nothing the courts do is going to stop that.

This decision, written by Chief Justice John G. Roberts Jr., only confirms that. His opinion didn’t overturn the rule that says that any time the government wants information on you that you’ve provided to a third party, the government can get it, even though in the digital age most of your information is now in the hands of third parties like cellphone companies, internet providers and social-media giants like Facebook. He stressed how “narrow” the decision was and didn’t rule out giving the government freedom to track you for up to six days.

Worse yet, if the court loses one of its liberal justices, this decision could be reeled back in an instant, since it was the court’s four liberals who joined with the chief justice for the majority. Three of the four dissenting conservative justices, in an opinion by Justice Anthony Kennedy, even opined, remarkably, that “a person’s movements are not particularly private.”

And two of them would even have wound back the law by more than 50 years to say the Fourth Amendment no longer protects “reasonable expectations of privacy” but only property rights, leaving us without constitutional recourse for all of our digital information.

What we need is regulation, and fast, and it is going to have to come from the local, state and federal governments, not the courts. Even Justice Samuel Alito, in his dissent, understood that “legislation is much preferable.” He said that was “for many reasons, including the enormous complexity of the subject, the need to respond to rapidly changing technology, and the Fourth Amendment’s scope.”

Almost every day brings another story of how policing technology is chiseling away at our civil rights and liberties. License plate readers already record our movements on the ground; drones will soon peer down at us from the skies. Artificial intelligence will power facial recognition tools in CCTV and police body cameras to identify us as we move about. Algorithms will target where crime will supposedly occur next, and who might commit it. All of these technologies will be in the hands of government agencies that have the power to keep us under watch, to handcuff us, to lock us up.

There is promise in these technologies, to be sure. Drones can locate a missing child or map a traffic accident from the sky so that cars and trucks aren’t stuck for hours. Pod cameras — short for police observation devices, affixed to utility or traffic poles or buildings — can catch criminals in the act in violence-torn neighborhoods. Location tracking of cellphones can identify serial armed robbers.

For all of its benefits, however, policing technology can also do unfathomable damage.

Let’s start with the obvious: It will cut deeply into our privacy. It was designed to do just that. We’re told technology will pick out the bad guy from a crowd, but if you’re in the crowd too, you’ll be observed. We’ll all be under the government’s watchful eye.

But the impact inevitably is going to fall most heavily on minorities and marginalized communities.

Some of the technology has racial injustice baked into it. Algorithms don’t have to look at race to be racist. Whether written by humans or a product of machine learning, algorithms take past facts and magnify them into future police actions. They rely heavily on criminal records. Much of street policing in recent years — stop and frisk, marijuana enforcement, catching fare-beaters — has been deployed disproportionately against minorities and in poor neighborhoods. Police may “go where the crime is,” but because so much focus has been on low-level offenses in disadvantaged areas that are ignored elsewhere, these algorithms make it inevitable that the police will return to these places time and again.

So what’s to be done?

As Carpenter makes clear, looking to the courts won’t be helpful. With rare exception, the courts have waved a green flag at police surveillance. The Fourth Amendment prohibits “unreasonable searches and seizures,” but if a government action is not considered a search or seizure, there is no protection at all. The fight in Carpenter took place because the use of most of these new technologies isn’t considered a search. That’s because you have no “reasonable expectation of privacy” in public places. This makes you fair game for all kinds of surveillance, from facial recognition to CCTV cameras. And warrants and probable cause — the usual tools of courts — won’t work anyway when the goal is to keep us all under watch.

We need rules that maximize the use of the technology of which we approve but that sharply curtail the risks. In some cases, we may want to rule out the technology for law enforcement use altogether. But for technologies we might allow, here are some basic guidelines:

First, no policing technology should be used unless the public has a voice in the decision to authorize it.

Second, citizens can’t make intelligent decisions about policing technology unless they understand what the technology does and how it works. Police buy technology from private vendors that want to keep its details secret for proprietary reasons. That won’t fly when the technology is being used by our government on us.

Third, we need clear rules about deploying policing technology: when it can and can’t be used, how each use is authorized and by whom. Drones to map traffic accidents? Sure. To peer into windows of homes? No way. Surveillance devices to locate cellphones? Yes, but only with the appropriate court orders. Predictive policing to identify places where crime might occur next? Perhaps, but not if there’s a chance these algorithms will perpetuate the racially disparate impact of policing.

Officers using these technologies need proper training, and law enforcement agencies should not be allowed to use them unless they can show they understand what they are buying.

Fourth, if data is being collected, sharp, clear limits on its use must be defined. How long can this data be retained? Can it be shared with other agencies? Or combined with other data purchased from private vendors?

Finally, strong oversight is critical. The police must be closely monitored in their use of these technologies, and data on their performance must be disclosed regularly.

There are powerfully hard questions here. But courts aren’t going to answer them, and we can’t continue to duck them. We need to act now, because the technology is getting away from us, fast, and the unregulated harms could be devastating.

Barry Friedman is a law professor and director of the Policing Project at New York University School of Law. He is the author of “Unwarranted: Policing Without Permission.”