What is a Reasonable Expectation of Privacy in the Digital Age?

SCOTUS takes up geofence warrants and the Fourth Amendment

Stay Tuned is going live! Preet Bharara and Barb McQuade will be at The 92nd Street Y in New York City on May 31 to discuss her new book, The Fix: Saving America from the Corruption of a Mob-Style Government. Get tickets here.

---

(Photographer: Graeme Sloan/Bloomberg via Getty Images)

As technology rapidly evolves, the law limps along behind to keep up.

On Monday, the Supreme Court heard arguments pertaining to geofence warrants, which investigators routinely use to reverse-identify suspects at crime scenes. In Chatrie v. United States, the defendant was convicted based on cell phone location evidence that placed him at a bank at the time of an armed robbery. Chatrie argues that the geofence warrant used to identify him violated his Fourth Amendment right against unreasonable searches.

The Fourth Amendment protects “[t]he right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures,” and provides that “no warrants shall issue, but upon probable cause, supported by oath or affirmation, and particularly describing the place to be searched, and the persons or things to be seized.” As technologies become more sophisticated and information proliferates, law enforcement and the courts are reckoning with what exactly constitutes a search and how far individuals’ right to privacy extends.

Geofence warrants, often described as “reverse search warrants,” require companies to provide information that can identify the smartphones present in a particular location at a particular time – typically, at the place and time of a given crime. Investigators then use that information to reverse-identify possible suspects.

In this case, a police detective obtained a warrant from a state court directing Google to produce the location records of cell phone users in a three-step process, allowing investigators to progressively narrow the list of suspected individuals through their devices. The first step produced a list of 19 devices in the area of the bank at the time of the robbery, but without the names of the users. Step two, with different parameters, yielded a subset of nine devices. In the third step, a further narrowed set of criteria listed three devices. As directed by the warrant, only then did Google unmask the names of the users, including Chatrie. Investigators used this information to obtain a search warrant for Chatrie’s home, which contained robbery demand notes and $100,000 in cash, wrapped in bands signed by the victim teller. Facing this evidence, Chatrie entered a conditional guilty plea, allowing him to preserve the Fourth Amendment issue for appeal. The Fourth Circuit held that the execution of the warrant did not constitute a “search” under the Fourth Amendment, because Chatrie had no “reasonable expectation of privacy in two hours’ worth of Location History data voluntarily exposed to Google.”

Although Google itself no longer keeps location information, the justices remain interested in the legality of geofence warrants, which are still used to seek location data from other companies, such as Apple, Uber, Lyft, and Snapchat.

In recent years, the Court has restricted searches in cases involving new technologies, requiring a warrant to install GPS trackers on vehicles, review the contents of cell phones, or obtain cell phone users’ “cell site” location information (automatically generated location information based on cell users’ proximity to cell towers). The latter case, Carpenter v. United States, arose in 2011, when I was serving as U.S. Attorney in Detroit.

Based on a cooperator’s information that Timothy Carpenter was part of an armed robbery crew, prosecutors in my office sought location information from cell phone providers to corroborate his presence at the crime scenes. At the time, courts permitted investigators to obtain cell site location information with a court order under the Stored Communications Act based on “specific and articulable facts showing that there are reasonable grounds to believe” that the records sought “are relevant and material to an ongoing criminal investigation,” a standard lower than probable cause. The records put Carpenter at the scene of seven armed robberies, and he was convicted at trial based in part on the cell site location evidence. The case went to the Supreme Court, which ruled that when seven days of cell site location information is collected, the government must obtain a search warrant, which requires the higher showing of probable cause to believe a crime has been committed and that evidence of the crime will be found in the information sought.

The Court’s Carpenter opinion, authored by Chief Justice John Roberts, focused on two aspects of data privacy involving cell phones. The first concern was the involuntary surrender of location data. Because a cell phone is an essential tool of modern life, Roberts wrote, virtually all of us are constantly sharing our location information without our affirmative consent. The second privacy concern was the nature of the information, which could provide a near-perfect, historical 24/7 record of one’s location. For these reasons, the Court concluded that the use of these records without a warrant violated Carpenter’s reasonable expectation of privacy. His conviction was nevertheless affirmed under the good faith exception that allows the use of evidence if it was obtained with the belief that its collection complied with the law. (We likely could have obtained a warrant had we known the Court’s view.)

While there will no doubt be some cases where investigators seeking cell site location information will be unable to satisfy the Fourth Amendment’s probable cause standard, and the information will therefore be inaccessible to them, that is the cost of living in a society that values privacy and civil liberties.

But the use of geofence warrants is different from the collection of an individual’s cell site location information at issue in Carpenter in several important ways. First, a warrant is obtained, even though the Court has never specifically held that one is required. In fact, Google typically will not produce the location information without a warrant. Second, in contrast to Carpenter, the surrender of private data here is voluntary. Carpenter involved automatically generated cell site location information held by the defendant’s phone company. By contrast, geofence warrants use opt-in geolocation data – that is, Google collects this location information only once an individual user has affirmatively agreed to share it (as opposed to phone companies’ auto-collection of cell site location information). According to the government’s brief, only about one-third of consumers do so. Third, the data obtained with a geofence warrant is limited to a finite duration around the time of the bank robbery, as authorized by the warrant, instead of the 24/7 chronicle of private life implicated in Carpenter.

In my view, the most significant concern raised by Chatrie is that the geofence warrant fails the particularity requirement of the Fourth Amendment (that warrants must “particularly describ[e] the place to be searched, and the persons or things to be seized”), improperly enabling the government to “search first and develop suspicions later,” and collecting information in an overly broad manner. The records obtained reveal not only the phone number that belongs to the suspect, but also the phone number of anyone in the vicinity. Of course, incidental collection of innocent people occurs in other contexts, such as when an authorized wiretap collects the conversations of individuals who participate in phone calls with the target. Such incidental collection, which is minimized by stopping recordings and destroying records, is often accepted as a collateral consequence of authorized collection. Moreover, in the case of geofence warrants, the data of innocent individuals is largely protected by the anonymization process.

Based on the justices’ questions at oral argument, a majority seems skeptical of Chatrie’s arguments but may be willing to place some limits on the use of geofence warrants. First, it seems likely that the Court will conclude that geofencing is a search that requires a warrant. This will simply ratify the current practice. Second, the Court may require the government to go back to the judge at each step with the information obtained in the prior phase to obtain new authorization, providing oversight over how investigators pare down their lists. Third, the Court could require narrower scopes for the radius and duration of the collection. Such modest fixes could preserve the tool for law enforcement while protecting the privacy of innocent cell phone users.

Certain justices also cautioned that the government’s arguments could implicate ever more data – individuals’ private photos, messages, and documents. Or that such warrants could reverse-identify individuals at sensitive locations, like houses of worship or other organizations. Perhaps most critically, the justices grappled with what this case entails for individuals’ reasonable expectations of privacy in the digital age.

One additional concern is whether we should trust this tool in the hands of Trump’s DOJ. With the president declaring himself the nation’s chief law enforcement officer, hell-bent on retribution, we should be vigilant in regulating executive power. But, of course, the Court’s decision will also apply to future administrations, as well as to state and local law enforcement agencies working to solve crimes. After all, geofence warrants were used to investigate and identify many of the January 6 attackers of the U.S. Capitol. But to find geofence warrants unconstitutional would sacrifice an important tool for all law enforcement agents, even as we enter uncharted technological territory.