Geofence Warrants – Part 3: The Grand Canyon Split
Two federal circuits. The same technology. The same constitutional amendment. Opposite conclusions. One commentator called it “the Grand Canyon of circuit splits”—but both circuits agree on the fundamentals. The disagreement is about the shape of Fourth Amendment protection, not its existence.
—
On August 9, 2024, three weeks after the Fourth Circuit’s en banc decision in Chatrie, the Fifth Circuit issued its ruling in United States v. Smith.[1] The case was nearly identical to Chatrie: a robbery, a geofence warrant, a defendant convicted on the basis of location data obtained from Google. The same technology, the same process, the same three steps.
The Fifth Circuit held that geofence warrants are unconstitutional.[2]
Not merely that they require a more particularized showing, or that they need additional safeguards. The Fifth Circuit held that geofence warrants are categorically unconstitutional—modern-day general warrants that the Fourth Amendment does not permit, even with probable cause.[3]
This was the opposite of the Fourth Circuit’s conclusion, and it created what one privacy advocacy group dubbed “the Grand Canyon of circuit splits.”[4] But the label deserves scrutiny. Both circuits agree that the Fourth Amendment applies to geofence warrants—both require a warrant and probable cause. The disagreement is about what the warrant must describe: whether the Fourth Amendment demands individualized particularity (the Fifth Circuit’s position) or whether a warrant describing a geographic area and time window is sufficient (the Fourth Circuit’s position). This is a disagreement about the shape of Fourth Amendment protection, not about whether protection exists at all.
The divergence between the Fourth and Fifth Circuits reveals a structural problem that existing doctrine cannot resolve—but the problem is more nuanced than a simple binary suggests.
—
How Both Circuits Applied Carpenter—And Reached Opposite Conclusions
Both circuits began their analysis with Carpenter v. United States (2018), the Supreme Court’s most recent major Fourth Amendment decision.[5] Carpenter held that the government’s acquisition of cell-site location information (CSLI) from wireless carriers constitutes a Fourth Amendment search, requiring a warrant. Chief Justice Roberts’s majority opinion rested on two key observations: that CSLI provides a “detailed, encyclopedic, and effortlessly compiled” record of a person’s movements, and that CSLI is collected passively by cell phone companies without any affirmative act by the user.[6]
The Fourth Circuit read Carpenter narrowly. Because Carpenter distinguished CSLI on the ground that it is collected passively, the Fourth Circuit held that Location History data—actively enabled by users who opt into Google’s services—is outside Carpenter‘s protection and governed by the traditional third-party doctrine from Smith v. Maryland.[7] Under the third-party doctrine, when you share information with a third party, you assume the risk that the third party will disclose it to the government. Result: no Fourth Amendment search.[8]
The Fifth Circuit read Carpenter broadly. Carpenter, the court held, established that “detailed, encyclopedic” location records are constitutionally protected regardless of how they are collected. The distinction between passive collection (CSLI) and active opt-in (Location History) is constitutionally irrelevant when the data is equally revealing.[9] A Google Location History log, updated every two minutes, provides a more detailed record of a person’s movements than CSLI—which records only the cell tower nearest to a phone at a given moment, not the phone’s precise GPS coordinates. If the less precise data in Carpenter is protected, the more precise data in a geofence warrant must be too.
Both readings are plausible. That is the problem. The Fourth Amendment’s text and history do not clearly dictate which reading is correct. Carpenter established a principle—that detailed digital location records can be constitutionally protected even when held by a third party—but left the scope of that principle deliberately vague. The circuits’ disagreement is not about Carpenter‘s holding but about its logic: does Carpenter protect location data because of its precision and comprehensiveness (the Fifth Circuit’s reading), or does it protect only data that is collected passively without affirmative user action (the Fourth Circuit’s reading)? The answer will determine whether the Carpenter exception to the third-party doctrine is narrow or broad—and whether geofence warrants survive.
—
Voluntary Disclosure: A Doctrine Under Strain
The deepest disagreement between the circuits concerns the meaning of “voluntary disclosure” in the digital age.
The Fourth Circuit treated Google’s Location History opt-in as a meaningful choice. Users who enable Location History, the plurality held, have voluntarily shared their location data with Google and therefore have no reasonable expectation of privacy in that data.[10] Under the traditional third-party doctrine, when you tell Google where you are, you accept the risk that Google will tell the government.
The Fifth Circuit rejected this framing entirely. “Opting” to enable location tracking, the court held, is not a meaningful choice when carrying a smartphone with location capabilities is practically mandatory for participating in modern life.[11] The Fifth Circuit cited the same internal Google email that described the opt-out interface as “difficult enough that people won’t figure it out”[12]—evidence that even Google understood the “voluntary” nature of location sharing to be something of a fiction.
The Fifth Circuit also invoked a society-level assessment that the Fourth Circuit explicitly rejected. The Fourth Circuit focused on the individual: what did Chatrie choose to share? The Fifth Circuit focused on the system: what does geofence surveillance look like when deployed at scale? “Perhaps the most alarming aspect of geofences,” the Fifth Circuit wrote, “is the potential for ‘permeating police surveillance'”—”near perfect surveillance” at that.[13]
This disagreement tracks a broader tension in Fourth Amendment jurisprudence—one that Carpenter itself highlighted. Carpenter recognized that the third-party doctrine, as traditionally applied, is a poor fit for the digital age. The majority carved an exception for CSLI, holding that “the Government’s acquisition of [cell-site location information] constitutes a Fourth Amendment search” despite the data being held by a third party.[14a] Justices Thomas and Gorsuch went further in their separate writings, signaling willingness to reconsider the doctrine entirely.[14b]
The “reasonable expectation of privacy” test from Katz v. United States (1967) has always contained two inquiries: a subjective one (did this person expect privacy?) and an objective one (is that expectation one that society is prepared to recognize as reasonable?).[14] The Fourth Circuit leaned heavily on the subjective inquiry—Chatrie opted in, so he assumed the risk. The Fifth Circuit leaned heavily on the objective inquiry—society cannot recognize as reasonable a system that enables dragnet surveillance of everyone near any crime scene. But the deeper problem is that Carpenter already demonstrated the third-party doctrine’s instability: if CSLI collected passively warrants an exception, what principled line keeps actively-shared but equally-revealing Location History data on the other side?
—
Particularity: The Database Too Big to Search
The Fifth Circuit’s second holding—more dramatic than the first—concerns the particularity requirement. The court held that the Sensorvault database is simply too large to be searched constitutionally.[15]
The Fourth Amendment requires that warrants “particularly describ[e] the place to be searched, and the persons or things to be seized.”[16] Traditional warrants satisfy particularity by identifying a specific place (a house, an office, a car) to be searched. Geofence warrants, the Fifth Circuit held, violate particularity at the first step: the “place to be searched” is not a house or a car—it is a database of 592 million accounts.[17]
The court’s analysis was blunt: “law enforcement cannot obtain its requested location data unless Google searches through the entirety of its Sensorvault—all 592 million individual accounts—for all of their locations at a given point in time. Moreover, this search is occurring while law enforcement officials have no idea who they are looking for, or whether the search will even turn up a result.”[18]
This is a constitutional claim with far-reaching implications. If a database is too large to search constitutionally, then the size of the database itself becomes a constitutional fact. The more data a company collects, the more constitutional protection that data receives—not because the data is more sensitive, but because searching it is more invasive.
Professor Orin Kerr, writing on the Volokh Conspiracy, challenged this reasoning. Kerr argued that the Fifth Circuit’s particularity analysis confuses the “place to be searched” with the database itself, when the constitutional question should focus on the records the government actually obtains.[19] Under Kerr’s framework, a geofence warrant that returns only the records of devices in a specific area during a specific time is no less particular than a warrant to search a specific apartment in a large building. The building (or database) is large, but the search is narrow.
Kerr’s amicus brief, filed with the Supreme Court on April 2, 2026, elaborated this position: “The Fourth Amendment does not present an all or nothing choice between zero protection and absolute protection. Where the law requires a warrant, it also provides a means to draft a lawful warrant.”[19a] Kerr argued that a “properly drawn geofence warrant can satisfy the Fourth Amendment”—narrow in both time and space at Step 1, with the particularity concern focused on Step 2 (the identification of specific devices from the anonymized list). This middle position—neither banning geofence warrants nor giving them a blank check—represents a third path the Court could take.
The Fifth Circuit was unmoved. Google Maps creator Brian McClendon’s description of geofence warrants as “fishing expeditions”[20]—quoted in the opinion—captured the court’s concern: the Framers prohibited general warrants precisely because they authorized searches of large collections of people and places without particularized suspicion. The size of the collection being searched is not irrelevant to the constitutional analysis—it is the constitutional analysis.
—
The Harvard Law Review Diagnosis: Probable Cause Can’t Solve This
The Harvard Law Review, in its February 2025 analysis “Much Ado About Geofence Warrants,” identified the structural problem that neither circuit fully addressed: probable cause—the only readily available standard to filter warrant applications—can nearly always be shown in the geofence context.[21]
When a database contains the location histories of 592 million people, probable cause that someone near a crime scene has committed a crime is almost always available. The standard was designed for a world where the government had to identify a suspect before searching. In the geofence context, the government searches first and identifies the suspect second. Probable cause that the search will reveal evidence of a crime somewhere in the database is not the same as probable cause that a specific person committed a crime—but current doctrine provides no clear way to distinguish the two.
Professor Paul Ohm identified this problem in 2010: “In increasingly common situations, whenever police have any suspicion at all about a piece of evidence, they almost always have probable cause.”[22] Geofence warrants are the technological realization of Ohm’s warning. When probable cause is this easy to satisfy, it stops functioning as a meaningful constitutional limit.
The Harvard Law Review analysis concluded that a theoretical middle ground—allowing geofence warrants with a probable cause requirement—would “collapse into the Fourth Circuit’s unqualified authorization,” because probable cause can be shown in virtually any geofence case.[23] This is why the Fifth Circuit held geofence warrants categorically unconstitutional: not because they could be used responsibly, but because existing constitutional standards provide no administrable way to ensure they are used responsibly.
But the Harvard Law Review analysis also suggested that the problem may be solvable through doctrinal innovation rather than categorical prohibition. If the Court were to develop a framework that distinguishes between narrow geofence warrants (small area, short time, serious crime) and broad ones (neighborhood-scale, days-long, routine investigation), the probable cause problem could be addressed through scope-based limits rather than a blanket ban. Professor Kerr’s amicus brief, filed after the Harvard Law Review analysis, proposes exactly this kind of framework: a properly drawn geofence warrant, narrow in both time and space, can satisfy the Fourth Amendment’s particularity requirement.[23a] The question is whether the Court is willing to engage in that kind of case-by-case particularity analysis, or whether it prefers a clear rule—either permitting or prohibiting geofence warrants categorically.
The Supreme Court will have to decide which circuit’s reading of the Fourth Amendment is correct—or, perhaps, whether the Fourth Amendment as currently understood is adequate to address mass digital surveillance at all. The scale of the problem—11,500+ warrants per year, innocent bystanders routinely identified, no statutory limits—makes that decision urgent. That is the story of Part 4.
—
Notes
-
*United States v. Smith*, 110 F.4th 817 (5th Cir. 2024). The opinion is available at https://www.ca5.uscourts.gov/opinions/pub/23/23-60321-CR0.pdf. ↩
-
Id. at 831 (holding that “geofence warrants present the exact sort of ‘general, exploratory rummaging’ that the Fourth Amendment was designed to prevent”). ↩
-
Id. (holding that the Fourth Amendment does not countenance geofence warrants “notwithstanding probable cause that evidence would be found in the searched records”). ↩
-
“Geofence Warrants and the Grand Canyon of Circuit Splits,” Protect Privacy Now, https://www.protectprivacynow.org/news/geofence-warrants-and-the-grand-canyon-of-circuit-splits. ↩
-
*Carpenter v. United States*, 585 U.S. ___, 138 S. Ct. 2206 (2018). ↩
-
Id. at 2216-17 (describing CSLI as “detailed, encyclopedic, and effortlessly compiled” and noting that CSLI is collected “passively” without affirmative user action). ↩
-
*United States v. Chatrie*, 107 F.4th 319, 338-40 (4th Cir. 2024) (plurality opinion distinguishing *Carpenter* on the ground that Location History data is actively shared by users); see also *Smith v. Maryland*, 442 U.S. 735 (1979). ↩
-
*Smith v. Maryland*, 442 U.S. 735, 743-44 (1979) (holding that “a person has no legitimate expectation of privacy in information he voluntarily turns over to third parties”). ↩
-
*United States v. Smith*, 110 F.4th 817, 826-27 (5th Cir. 2024) (rejecting the distinction between passive collection and active opt-in as constitutionally irrelevant when the data is equally or more revealing). ↩
-
*United States v. Chatrie*, 107 F.4th 319, 338-40 (4th Cir. 2024) (plurality opinion: “Just like in [Smith and Miller], Chatrie volunteered incriminating information about himself to a third party.”). ↩
-
*United States v. Smith*, 110 F.4th 817, 827-28 (5th Cir. 2024) (holding that opting into location tracking is a practical necessity, not a voluntary choice, in modern life). ↩
-
Id. at 828 (citing internal Google email disclosed in litigation); see also Amster & Diehl, “Geofence Warrants and the Fourth Amendment,” 74 *Stan. L. Rev.* 385, 410 (2022), https://review.law.stanford.edu/wp-content/uploads/sites/3/2022/02/Amster-Diehl-74-Stan.-L.-Rev.-385.pdf. ↩
-
*United States v. Smith*, 110 F.4th 817, 828 (5th Cir. 2024) (quoting the *Chatrie* dissent and characterizing geofence surveillance as “permeating police surveillance”). ↩
-
*Katz v. United States*, 389 U.S. 347, 361 (1967) (Harlan, J., concurring) (establishing the two-part test: subjective expectation of privacy that society recognizes as reasonable). ↩
-
*United States v. Smith*, 110 F.4th 817, 831 (5th Cir. 2024) (holding that the Sensorvault database is too large to satisfy the Fourth Amendment’s particularity requirement). ↩
-
U.S. Const. amend. IV. ↩
-
*United States v. Smith*, 110 F.4th 817, 831 (5th Cir. 2024). ↩
-
Id. ↩
-
Orin Kerr, “The ACLU’s Response to My Post on the Fifth Circuit’s Smith Ruling—and My Reply to the ACLU,” Volokh Conspiracy, *Reason* (Aug. 16, 2024), https://reason.com/volokh/2024/08/16/the-aclus-response-to-my-post-on-the-fifth-circuits-smith-ruling-and-my-reply-to-the-aclu/; see also Orin Kerr, “The Fifth Circuit Shuts Down Geofence Warrants—and Maybe a Lot More,” Lawfare (Aug. 13, 2024), https://www.lawfaremedia.org/article/the-fifth-circuit-shuts-down-geofence-warrants-and-maybe-a-lot-more. ↩
-
Jennifer Valentino-DeVries, “Tracking Phones, Google Is a Dragnet for the Police,” *New York Times* (Apr. 13, 2019), https://www.nytimes.com/2019/04/13/technology/google-sensorvault-location-tracking.html (quoting Brian McClendon). ↩
-
“Much Ado About Geofence Warrants,” *Harvard Law Review* (Feb. 2025), https://harvardlawreview.org/blog/2025/02/much-ado-about-geofence-warrants/ (“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.”). ↩
-
Paul Ohm, “The Probable Cause Mismatch,” 94 *Minn. L. Rev.* 521, 523 (2010), https://www.minnesotalawreview.org/wp-content/uploads/2012/03/Ohm_MLR.pdf. ↩
-
“Much Ado About Geofence Warrants,” *Harvard Law Review* (Feb. 2025), https://harvardlawreview.org/blog/2025/02/much-ado-about-geofence-warrants/. ↩
-
*Carpenter v. United States*, 585 U.S. ___, 138 S. Ct. 2206, 2213 (2018) (recognizing an exception to the third-party doctrine for cell-site location information). ↩
-
*Carpenter*, 138 S. Ct. at 2262 (Thomas, J., concurring) (arguing for reconsideration of the third-party doctrine); *Carpenter*, 138 S. Ct. at 2273 (Gorsuch, J., dissenting) (suggesting property-based approaches). ↩
-
Orin Kerr, “My Amicus Brief in the Geofence Warrant Case, United States v. Chatrie,” Volokh Conspiracy, *Reason* (Apr. 2, 2026), https://reason.com/volokh/2026/04/02/my-amicus-brief-in-the-geofence-warrant-case-united-states-v-chatrie/; see also Brief of Professor Orin S. Kerr as Amicus Curiae in *Chatrie v. United States*, No. 25-112 (U.S. Apr. 1, 2026), https://www.supremecourt.gov/DocketPDF/25/25-112/403349/20260401133909957_25-112acProfessorOrinSKerr.pdf. ↩
-
Brief of Professor Orin S. Kerr as Amicus Curiae in *Chatrie v. United States*, No. 25-112 (U.S. Apr. 1, 2026), https://www.supremecourt.gov/DocketPDF/25/25-112/403349/20260401133909957_25-112acProfessorOrinSKerr.pdf (arguing that “a properly drawn geofence warrant can satisfy the Fourth Amendment” with particularity focused on what the government actually obtains). ↩