In 2020, law enforcement agencies served Google with more than 11,500 warrants directing the company to conduct sweeping searches of its vast location database. These “geofence warrants” instructed Google to search detailed location information logged by hundreds of millions of devices and to return lists of users found to be within a particular area during a specified time period. Geofence warrants have been used to identify the perpetrators of serious crimes. They have also turned innocent passersby — a man tracking his bike rides in the vicinity of a burglarized home, for example — into suspects. Recently, the Fourth and Fifth Circuits issued diametrically opposed rulings on the constitutionality of geofence warrants. This development dispelled any lingering delusions that current Fourth Amendment doctrine is a useful and coherent fit for private mass-surveillance technology. In particular, geofence warrants and other mass-surveillance tools have obliterated the utility of probable cause as a meaningful standard for balancing privacy interests against law enforcement needs.
United States v. Chatrie (4th Cir. 2024) and United States v. Smith (5th Cir. 2024) began as near-twins in relevant background and posture. In both cases, the defendants were convicted of robbery after the respective district courts denied their motions to suppress evidence obtained by way of geofence warrants. The defendants challenged the introduction of geofence-derived evidence on appeal. Both circuits were called upon to decide the constitutionality of the process by which law enforcement elicits eventually de-anonymized location information from private companies like Google.
That exhausts the similarities. The Fifth Circuit found a reasonable expectation of privacy in location history data where the Fourth Circuit found none. The Fourth Circuit viewed the need to “opt in” to location data-sharing as ensuring knowing and voluntary disclosure, where the Fifth Circuit saw “opting” to enable location tracking as a practical necessity in the digital age. The Fourth Circuit hung its decision on the individual-level output of a geofence search, which it called a “brief glimpse into [someone’s] whereabouts” unlikely to “offer insight into his habits, routines, and associations.” The Fifth Circuit not only rejected that characterization, noting that “the potential intrusiveness of even a snapshot of precise location data should not be understated,” but also anchored its reasoning in a society-level assessment: “Perhaps the most alarming aspect of geofences is the potential for ‘permeating police surveillance’”— “near perfect surveillance” at that.
Ultimately, the Fourth Circuit held that a geofence warrant yielding two hours’ worth of precise location data involves no Fourth Amendment search and thus need not be supported by probable cause. The Fifth Circuit held not only that the practice constitutes a Fourth Amendment search but also that, given the massive scale of the database at issue, the Fourth Amendment does not countenance geofence warrants at all, notwithstanding probable cause that evidence would be found in the searched records.
Much has been made of the Fourth and Fifth Circuits’ divergence, which one commentator dubbed “the Grand Canyon of circuit splits.” Less discussed is that the disparity is the inevitable consequence of existing Fourth Amendment doctrine being a poor fit for big technology in general and private mass-surveillance technology in particular.
Current Fourth Amendment doctrine provides no administrable way to mediate between allowing and banning all geofence warrants because probable cause — the only readily available standard to filter warrant applications — can nearly always be shown in the geofence context. That’s partly because of location databases’ massive scale. When Google’s location data repository (the “Sensorvault”) was fully operational, it contained detailed location information from devices attached to 592 million individual accounts. Location data was logged on average every two minutes from Android devices and Google apps on non-Android cell phones. Data-sharing is nominally but — as the Fifth Circuit noted — not realistically opt-in. (An internal Google email explained that the interface “feels like it is designed to make [opting out] possible” but is “difficult enough that people won’t figure it out.”) Add to the mix that a cell phone with location tracking capabilities is so ubiquitous as to be practically “a feature of human anatomy” and you get probable cause that a geofence search will turn up evidence of a crime on basically any scene.
Geofences’ resistance to probable cause–based regulation rings of Professor Paul Ohm’s observation in 2010, amid rapid change in communications and surveillance technologies, that “[i]n increasingly common situations, whenever police have any suspicion at all about a piece of evidence, they almost always have probable cause.” Geofencing capabilities are no exception. A theoretical middle ground between the Fourth and Fifth Circuits’ rulings — allowing geofence warrants on the readily satisfied condition of probable cause — would collapse into the Fourth Circuit’s unqualified authorization.
Professor Orin Kerr has raised this all roads lead to Rome–style argument about even the Fifth Circuit’s hardline anti-geofence ruling; that is, the Fifth Circuit’s ruling, despite appearing the diametric opposite of the Fourth Circuit’s, could ultimately also lead to little or no privacy protection. Kerr argues that permitting warrants for a type of information tends to extend Fourth Amendment protections vis-à-vis that type of information. Take wiretaps, which Kerr uses as an historical example. Wiretapping was not treated as a search “at a time when no warrant could be obtained to engage in wiretapping.” But soon after warrants for wiretaps were permitted, courts started treating wiretaps as searches subject to Fourth Amendment limitations. On Kerr’s view, it doesn’t much matter whether a court treats geofencing as a non-search immune from the warrant requirement (as the Fourth Circuit did) or as a search that can’t be justified even by a valid warrant (as the Fifth Circuit did). When the pressure to geofence arises — when a serious crime looks unsolvable without a geofence search, perhaps — the Fifth Circuit’s position will make it “harder . . . to argue that a warrant is needed to conduct that kind of surveillance in the first place.” The Fifth Circuit’s hardline approach could in practice devolve into the Fourth Circuit’s lack of protection.
The Fourth Circuit has another chance to effect meaningful legal regulation of geofence warrants: It will soon rehear Chatrie en banc. The full court should heed the Fifth Circuit’s warning about the pervasive police surveillance that geofence warrants enable. It should also attend to Professor Kerr’s admonition that a blunt prohibition like the Fifth Circuit’s could ultimately prove powerless to rein in law enforcement’s capitalizing on private mass surveillance. It should consider creative ways to heighten probable cause requirements in the geofence context, accounting for the impossibility of accommodating mass surveillance databases with traditional probable cause limitations alone.
While courts wrestle with options for sophisticated legal regulation of geofences, innovation outside of the criminal procedure context may be a stopgap. Technology companies wield ultimate control over geofence warrants’ efficacy. Case in point: In 2023, while Chatrie and Smith were pending, Google announced that it would phase out the Sensorvault. If that happens, it will relieve significant pressure in the geofence warrant debates: While many private companies that track location data — including Lyft, Uber, Apple, and Snapchat — have received geofence warrants, Google is “the most common recipient and the only one known to respond.” Without means to compel Google or other private companies to retain or organize their data, law enforcement agencies are at the mercy of large private companies with respect to geofences. (The status of Google’s pronounced changes is unclear. And, where constitutional, warrants may still issue for data previously stored in the Sensorvault.) The possibility of tech companies’ self-regulation will be little comfort to people like Jorge Molina, who spent six days in jail and consequently lost his job after he forgot to sign out of Google before lending his phone to a man later arrested for murder. If current probable cause doctrine is ill-equipped to accommodate the advent of intimate location data logged and stored at a scale previously inconceivable, something must give — quickly, before the next big development in “creepy” mass location surveillance technology takes hold.
The post <strong>Much Ado About Geofence Warrants</strong> appeared first on Harvard Law Review.