Supreme Court Takes Up Police Use of Cell Phone Location Data in Criminal Cases

On a spring morning in 2019, a man walked into a Call Federal Credit Union in Midlothian, Virginia, slid a note to the teller demanding "at least 100k," and walked out with a stack of cash [1]. Police had no suspect. No eyewitness could identify the robber. But a detective had another idea: ask Google to identify every cellphone user who had been near the bank during the robbery.

That request — a "geofence warrant" — ultimately led investigators to Okello Chatrie, who was convicted and sentenced to nearly 12 years in federal prison [2]. On April 27, 2026, the Supreme Court heard oral arguments in Chatrie v. United States, a case that could redefine the boundaries of digital privacy under the Fourth Amendment [3].

What Is a Geofence Warrant?

A geofence warrant works by drawing a virtual boundary — a "geofence" — around a geographic area where a crime occurred. Law enforcement then obtains a court order compelling a technology company, most often Google, to search its vast database of user location records and identify every device that was within that boundary during a specified time window [4].

The process typically unfolds in three steps. First, the company provides anonymized data on all phones detected in the geofenced area. Second, investigators request expanded location data for a subset of those devices to narrow the pool. Third, the company unmasks the identities of the remaining users [5]. In the Chatrie case, Google initially identified 19 users near the credit union; the detective eventually narrowed the list to Chatrie, whose phone had Google's "Location History" feature switched on [1].

Critics describe this as a "reverse warrant" — instead of identifying a suspect and then seeking evidence, police cast a wide net and work backward from data to find a suspect [6].

The Explosion of Geofence Requests

The scale of geofence warrant use expanded dramatically before Google changed its data practices. Google reported receiving 982 geofence warrants in 2018, 8,396 in 2019, and 11,554 in 2020 [7]. By 2021, geofence requests accounted for more than 25% of all warrants Google received in the United States [7].

Source: Google Transparency Reports / Court Filings

Data as of Jan 1, 2024CSV

These numbers reflect only Google's data. Apple, which stores location data differently, reported receiving just 36 geofence warrants in 2022 and 26 in 2023, underscoring that Google's centrality to the practice stems from its extensive Location History database, internally known as "Sensorvault" [7].

In December 2023, Google announced that it would begin storing Location History data on users' devices rather than centrally, delete it by default after three months, and encrypt any cloud backups so that even Google cannot read them [8]. These changes may render future geofence warrants against Google far less productive — but they do not resolve the constitutional question for the thousands of cases already built on the old system, nor do they prevent the technique from being applied to other data holders.

Carrier Transparency: The Broader Data Pipeline

Geofence warrants represent only one slice of law enforcement's appetite for location data. The major wireless carriers — AT&T, Verizon, and T-Mobile — field tens of thousands of law enforcement demands annually for cell-site location information (CSLI), the records generated every time a phone connects to a cell tower.

Verizon reported receiving 19,886 warrants for location data in the first half of 2023 alone, along with 3,013 warrants or court orders for "cell tower dumps" — bulk downloads of every device that connected to a specific tower during a given interval [9]. In the second half of 2024, Verizon received 25,031 total warrants from U.S. law enforcement [9]. T-Mobile reported receiving 51,557 historic CSLI requests in 2016, a figure that preceded the Carpenter warrant requirement [10].

Source: Verizon Transparency Reports

Data as of Jun 1, 2025CSV

AT&T publishes similar transparency data, breaking out requests for location information, emergency disclosures, and international demands [11]. The cost to agencies varies by carrier and request type, but the financial barrier is low enough that location data requests have become routine across federal, state, and local law enforcement.

The Constitutional Question: From Carpenter to Chatrie

The legal backdrop for Chatrie is the Supreme Court's 2018 decision in Carpenter v. United States, which held 5-4 that obtaining seven days or more of historical CSLI constitutes a "search" under the Fourth Amendment and requires a warrant supported by probable cause [12]. Before Carpenter, the government could access these records under the Stored Communications Act by showing merely "reasonable grounds" — a standard well below probable cause [12].

Chief Justice John Roberts, writing for the majority in Carpenter, rejected the application of the "third-party doctrine" to CSLI. That doctrine, rooted in Smith v. Maryland (1979), holds that individuals have no reasonable expectation of privacy in information they voluntarily share with third parties like banks or phone companies [13]. Roberts reasoned that CSLI is qualitatively different: it creates a "detailed, encyclopedic, and effortlessly compiled" record of a person's movements that is generated automatically, not truly "voluntarily" shared [12].

But the Carpenter majority deliberately left several questions open. The opinion explicitly declined to address real-time CSLI, tower dumps, or "other collection techniques" [12]. Chatrie sits squarely in that gap.

The Doctrinal Fault Lines

The Supreme Court is being asked to resolve which Fourth Amendment framework governs geofence warrants. Three competing doctrines are in play.

The third-party doctrine. The government argues that Chatrie opted into Google's Location History feature, voluntarily disclosing his location data to a third party, and therefore lacks a reasonable expectation of privacy under Smith v. Maryland [14]. If the Court adopts this view, geofence warrants would require no warrant at all — a significant expansion of the third-party doctrine into the digital era.

The mosaic theory. Under this framework, endorsed by several concurring justices in United States v. Jones (2012) and implicitly embraced in Carpenter, individual data points may not constitute a search, but their aggregation into a comprehensive picture of a person's life does [15]. Chatrie's attorneys argue that Google Location History — which can pinpoint a user within roughly three meters, log data every two minutes, record elevation within a building, and retain information indefinitely — is even more revealing than the CSLI at issue in Carpenter [5].

A new digital-privacy framework. Some amici have urged the Court to move beyond both doctrines and establish a clear rule that reverse warrants — those directed at identifying unknown suspects from a pool of data rather than gathering evidence about a known suspect — are categorically unconstitutional general warrants [16]. The ACLU and Electronic Frontier Foundation filed a brief asserting that "geofence warrants are never a permissible investigatory method under the Fourth Amendment" [16].

The en banc Fourth Circuit was deadlocked. Its 14 judges produced nine separate opinions and split 7-7 on whether Google's compliance with the geofence warrant even constituted a "search" [5].

The Law Enforcement Case

Law enforcement agencies and 18 states filed amicus briefs defending geofence warrants as a necessary tool for solving serious crimes [16]. They argue that the warrants should not be treated as categorically unconstitutional and that the relevant inquiry should focus on traditional warrant requirements — probable cause, particularity, and reasonableness [16].

The practical record supports their position in specific cases. Geofence warrants have been used to investigate murders, arsons, burglaries, sexual assaults, kidnappings, and terrorism, including the identification of participants in the January 6, 2021, Capitol breach [2]. In Louisville, Kentucky, police turned to geofence warrants to investigate homicides in a city struggling with rising violent crime [17]. A 2017 Milwaukee kidnapping and sexual assault — one of the earliest known uses of the technique — was solved after investigators obtained geofence data [2].

Justice Department lawyers told the Court that police had "probable cause" to believe Google possessed information that could identify the robber, accomplices, and witnesses, and that Chatrie's decision to enable Location History stripped him of any privacy expectation [1].

The Privacy Advocates' Response

Privacy organizations counter that the utility of geofence warrants in individual cases does not justify their sweeping scope. The Brennan Center for Justice, the Center for Democracy and Technology, and other groups argued in an amicus brief that geofence warrants are inherently overbroad because they necessarily capture data on hundreds or thousands of innocent people to identify a single suspect [16].

Justice Sonia Sotomayor flagged this concern during oral argument, noting that "data will follow you everywhere" and could reveal whether someone visited "a brothel or a cannabis store" [1].

Chatrie's attorney, Adam Unikowsky, told the justices: "The technology may be novel, but the constitutional problem it presents is not" [1]. His brief draws a direct line from the 18th-century "writs of assistance" — the general warrants that helped spark the American Revolution — to modern geofence warrants, arguing that both grant the government open-ended authority to search without identifying a particular suspect [5].

On the question of how investigations would proceed under stricter rules, privacy advocates point to traditional police work — witness canvassing, surveillance cameras, forensic evidence — and argue that the mere convenience of a digital shortcut does not justify abandoning constitutional protections [16]. They also note documented cases where CSLI evidence contributed to wrongful convictions: Vernon Horn and Marquis Jackson spent 18 years in prison before being exonerated with cell-site data that had not been disclosed to the defense, and Herbert Alford served five years on murder charges before new CSLI analysis cleared him [18].

The Real-Time vs. Historical Divide

Carpenter addressed historical CSLI — records generated over days or weeks and stored by carriers. But it explicitly reserved the question of real-time location tracking, and lower courts have since split on where the line falls.

The Fifth Circuit ruled in United States v. Smith (2024) that location data shared with Google was analogous to the CSLI in Carpenter and required a warrant based on probable cause and particularized suspicion, holding that the geofence warrant used in that case was constitutionally inadequate [19]. Other circuits have drawn different boundaries. No consistent time threshold separates "real-time" from "historical" across federal and state courts, creating a patchwork that Chatrie could resolve — or deepen.

International Comparisons

The United States is not alone in grappling with these questions, and international practice offers no simple answer.

European Union. The Court of Justice of the European Union ruled in La Quadrature du Net (2020) that EU member states cannot require telecom providers to carry out "general and indiscriminate" retention of traffic and location data [20]. Data retention must be targeted, limited to what is strictly necessary for combating serious crime, and subject to prior judicial review. Only a "genuine and present or foreseeable" threat to national security justifies a temporary exception [20]. Real-time location collection is permitted only for individuals suspected of terrorist involvement, and only with prior judicial authorization [20].

United Kingdom. The Investigatory Powers Act 2016 permits police and intelligence agencies to access communications data, including location records, but the Data Retention and Acquisition Regulations 2018 raised the threshold so that access is allowed only for "serious crime" — offenses carrying a sentence of 12 months or more [21]. A "double-lock" system requires both ministerial authorization and judicial approval for interception warrants [21].

Canada. The Supreme Court of Canada's 2014 decision in R v. Spencer established that internet users have a reasonable expectation of privacy in subscriber information held by service providers, and law enforcement must obtain prior judicial authorization before requesting such data [22]. The Court recognized "privacy as anonymity" as a constitutionally protected component of informational privacy [22].

Germany. German constitutional law imposes strict proportionality requirements on surveillance, and the Federal Constitutional Court has repeatedly struck down data retention laws as incompatible with the right to informational self-determination.

The EU framework is significantly more restrictive than current U.S. practice. Proponents of stricter warrant requirements cite international norms as evidence that effective policing does not require mass location data collection. Defenders of broader law enforcement access argue that the comparisons are inapt because U.S. crime rates and the scale of the federal system create different operational demands.

The Technology Gap

A ruling in Chatrie will address Google Location History data from 2019. But the technology landscape has already shifted.

Cell-site location information — the data at the center of Carpenter — is the least precise form of location tracking, typically accurate to 500 to 1,500 meters depending on tower density [23]. Google Location History, which combines GPS, Wi-Fi, and cell signals, can achieve accuracy within roughly three meters [5]. GPS alone is accurate to about 4.9 meters [23]. Wi-Fi triangulation can locate a device within 10 to 25 meters in dense urban environments [23]. Bluetooth beacons and ultra-wideband chips in newer devices can pinpoint location to within centimeters.

Each generation of technology is more precise, more persistent, and more capable of generating inferences about behavior. A ruling tailored to 2019-era Google Location History may not address the capabilities of 2026-era tracking, let alone what comes next. Several amici flagged this concern, urging the Court to articulate principles broad enough to govern future technologies rather than issuing a narrow, technology-specific holding [16].

Retroactive Exposure

If the Court rules that the geofence warrant in Chatrie was unconstitutional, the decision could ripple backward through thousands of cases. Google received more than 20,000 geofence warrants between 2018 and 2020 alone [7]. Many of those resulted in criminal prosecutions where defendants may not have raised Fourth Amendment challenges, either because the legal landscape was unsettled or because they lacked the resources to litigate the issue.

The scope of retroactive exposure depends on how broadly the Court rules and whether it applies the "good-faith exception" — the doctrine that evidence obtained by officers reasonably relying on existing law need not be suppressed even if the law later changes. The district court in Chatrie applied this exception, suppressing no evidence despite finding the warrant "plainly violates the rights enshrined in the [Fourth] Amendment" [5]. How the Supreme Court handles this question will determine whether a pro-defendant ruling has practical teeth or remains largely symbolic for past convictions.

What Comes Next

A decision is expected by the end of June 2026 [1]. The Court's options range from a narrow holding that the specific warrant in Chatrie was deficient to a broad pronouncement on whether reverse warrants are categorically incompatible with the Fourth Amendment.

Whatever the outcome, the case will not end the debate. Google's 2023 policy changes have already reduced the data available for geofence warrants. Law enforcement agencies are adapting, turning to other data sources — commercial data brokers, app-based location services, automated license plate readers — that may fall outside whatever rule the Court establishes [24]. Congress has considered but not passed legislation specifically addressing geofence warrants [6].

The 29 amicus briefs filed in Chatrie — from tech companies, law enforcement associations, civil liberties organizations, legal scholars, and state attorneys general — reflect the breadth of interests at stake [16]. The answer the Court delivers will shape not only the legality of a single Virginia bank robbery investigation but the constitutional framework for digital surveillance in the United States for years to come.