Virginia Court Blocks Geofence Warrant As Unconstitutionally Vague

Legal Issues

from the not-all-geowarrants-are-created-equal dept

There are multiple places government agencies can go to acquire location data. The Supreme Court’s Carpenter decision put a damper on warrantless demands, but the private sector has rushed to fill this hole in the law enforcement market by selling access to data pulled from apps that don’t (currently) require a warrant to access.

But this data tends to be vague and/or incomplete. There’s one company that collects location data with a frequency that rivals that of cell phone providers: Google. And while warrants may be technically in use, these warrants reverse the expectations of probable cause by turning everyone in a geofenced area into a suspect before investigators work backwards from the location data to generate a list of most likely suspects. Hence, the term “reverse warrants.”

Investigators have used these for years, flying under the radar while courts and citizens concerned themselves with bigger constitutional questions, like NSA surveillance and warrantless searches of arrestees’ cell phones. Courts are paying more attention now. And that has — in a few occasions — resulted in investigators having their geofence warrants rejected.

FourthAmendment.com brings us another rejection to add to the list of law enforcement failures — one that will contribute to the body of case law indicating that working backwards to probable cause isn’t how any of this works.

This decision [PDF] was handed down by the Fairfax County court in Virginia. The (poorly) redacted decision makes it clear law enforcement was in the wrong here — something made plainly apparent by this heading, which I will present as a silent movie title card for maximum comic effect:

If you can’t read/see/refuse to acknowledge my comic genius, the opinion subheading reads:

THE POLICE PRESENT A GEOFENCE WARRANT LACKING PROBABLE CAUSE AND PARTICULARITY.

The backstory is this: cops were investigating shootings at a motel in Fairfax County. Investigators apparently had no suspects so they asked Google to generate them a list of people who had been in the area at the time of the shooting based on their location data. Three “zones” were specified with two of them overlapping somewhat in hopes of catching the shooter who, according to CCTV footage from the motel, was shooting out of the rear of a moving car.

The investigators would get this list of “anonymized” numbers, and (as the county court notes) “without court involvement” review the list, and “without court involvement” return to Google to ask for an expanded search if no potential suspect is revealed by the first search. Finally (“without court involvement”), investigators would return to Google to ask for identifying data on those deemed close enough to call suspects.

This court isn’t willing to be cut off of nearly every step of this process, especially when it starts with a dearth of probable cause. The judiciary has a very important duty and this court isn’t going to ignore it. Things are moving quickly, but tech developments shouldn’t keep outpacing the public’s rights.

It happened relatively quickly, but ours is a highly surveilled nation. Video cameras seem to be everywhere, and data storage is practically unlimited. People cannot walk down the block in an urban area without being automatically recorded on multiple occasions. Almost everyone possesses a cell phone, that is now effectively a personal tracking device. When a crime occurs, police want to access this data to help them solve the crime. It is the duty of the judiciary to make sure the government’s use of new technology comports with familiar Fourth Amendment jurisprudence.

Warrants — no matter what they’re seeking — need certain things: particularity, probable cause, and guarantees from the government it won’t exceed the limitations of the Fourth Amendment. As the court points out, this geofence warrant has virtually none of these required ingredients.

Imagine if, in Ybarra, police knew someone in the bar possessed heroin, but they needed to identify who. Could a court authorize a warrant for police to search everyone in the bar to figure out who was the possessor? No. No court would properly authorize such a warrant. However, the Court sees little distinction between that search warrant application and the geofence search warrant application at issue in the present case. In both instances, police know a crime occurred but do not know the perpetrators. In both instances, police want to search everyone to find out who those people are. However, police may not do this physically in a bar full of patrons and a bartender. They similarly may not do so in a motel full of guests, visitors, and employees.

The Court finds there is no probable cause to search the motel patrons based on the present allegations. Without probable cause to conduct the search, the geofence search warrant must fail.

No probable cause. And no particularity. The police thought big, but that doesn’t help them here.

First, the police drew a GPS virtual zone that is geographically overbroad. It covers the entirety of the motel, the parking lot, and much of the residual property. […] Police have cast a net too broadly by seeking to search almost the entire motel property. Innocent motel patrons have a reasonable expectation of privacy within their respective rooms, as well as to common areas and the parking lot. There is no good reason to search patrons beyond the front area.

Second, the police seek data for too long a time. […] Here, police want just under 3 hours of data.. From surveillance video the police know the approximate time of the shooting—shortly after [redacted]. The shooting appears to have been a short duration event. By extending the time for which police seek data from Google to almost 3 hours, the police are likely to unnecessarily search too many motel patrons.

[…]

Third, by searching people in a motel and its grounds, police are targeting a particularly sensitive area. Motels are close proxies to one’s home on the scale of privacy expectations. In some ways, one’s privacy interests while in a motel exceed those in one’s own home. There are noble examples of this: a businessperson on a business trip wishing to not tip off a competitor as to an opportunity; or people gathering secretly for political purposes. There are ignoble purposes: a secret tryst. In both circumstances motel patrons expect privacy in their location data.

These facts become even more concerning when investigators — as they did in this case — believe they should be allowed to unilaterally alter the time and area covered by this already overly expansive request. Investigators’ belief they should be able to approach Google for identifying data without further submissions to the court help lead the court to reject this warrant — both in its current form and its proposed “no court needed” expansion.

For the time being, investigators have no warrant at all. The court has made it clear what isn’t acceptable. But given the location being searched by proxy with Google location data, it’s difficult to see how they’ll be able to work around the court’s restrictions. And this is a good thing. Probable cause is the baseline and geofence warrants don’t even try to approach that constitutional guideline. If the only assumption needed is that Google holds location records, every reverse warrant would be acceptable. But the Constitution does not work that way. Law enforcement either needs to do a whole lot better crafting these so-called warrants or, better yet, go back to the basics and start looking for suspects first, rather than trying to blunder their way into them by sifting through tons of unrelated data.

Filed Under: 4th amendment, general warrant, geofence warrant, location data, probable cause