Geofence Warrants · Location Data · Fourth Amendment
Chatrie v. United States
609 U.S. ___ (2026) · No. 25-112 · Decided June 29, 2026
A comprehensive legal monograph on the Supreme Court’s first geofence-warrant decision, the constitutional status of precise cell-phone location history, the limits of the third-party doctrine, and the unresolved questions of probable cause, particularity, good faith, and technology-assisted mass identification.
Executive Summary
Chatrie v. United States is the Supreme Court’s first case directly addressing a geofence warrant. It marks a significant development in the Court’s modern Fourth Amendment jurisprudence because it moves beyond the question whether the government may obtain the long-term location history of a known suspect and addresses a different investigative model: a reverse-location search that begins with a place and a time and uses a technology company’s location database to identify possible suspects.
The case arose from the May 20, 2019 robbery of the Call Federal Credit Union in Midlothian, Virginia. Investigators knew the time and place of the crime and had surveillance footage indicating that the robber carried a cell phone, but they did not know his identity. Detective David Hylton sought a geofence warrant requiring Google to search its Location History repository for devices located within a 150-meter radius of the credit union during a one-hour period surrounding the robbery. The warrant contemplated a multi-step process. Google first returned anonymized identifiers for responsive devices. Investigators then selected a subset for additional location information and eventually a smaller subset for identifying subscriber information. The process led police to Okello Chatrie.
The district court treated the warrant as constitutionally defective but declined to suppress because the officers had acted in objectively reasonable reliance on a warrant issued by a magistrate at a time when geofence law was unsettled. A divided Fourth Circuit panel affirmed on a broader ground, reasoning that Chatrie did not have a reasonable expectation of privacy in two hours of Location History that he had enabled Google to collect. The Fourth Circuit granted rehearing en banc and ultimately affirmed the district court in a one-sentence per curiam judgment, accompanied by extensive separate opinions exposing deep disagreement among the judges about the search question, the warrant’s constitutionality, and the proper role of good-faith analysis.
The Supreme Court vacated. Justice Kagan’s opinion for the Court held that law enforcement conducts a Fourth Amendment search when it acquires a person’s historical Google Location History. The Court concluded that the reasons it had identified in Carpenter v. United States for protecting historical cell-site location information applied “as well or better” to Google Location History. Location History could be more precise than conventional CSLI; it could reconstruct movement retrospectively and at scale; and it functioned in significant part as a personal record used by the account holder, not merely as an internal business record generated by a service provider.
The majority rejected the government’s proposed distinction between long-term and short-term location data. Carpenter had expressly reserved whether there might be some limited period of CSLI acquisition that would fall outside the Fourth Amendment. In Chatrie, the government asked the Court to recognize such a “grace period” and to place the two hours of data at issue within it. The Court declined. It reasoned that even a short period of precise location information can reveal highly private facts, including a person’s presence at a medical office, political event, religious service, intimate meeting, or other sensitive location. More fundamentally, the Court rejected the idea that Fourth Amendment protection switches on only after the government obtains a sufficiently large quantity of otherwise protected information.
The Court also rejected the government’s reliance on the third-party doctrine. The fact that Google stored the information did not place it outside the Fourth Amendment. Drawing on Carpenter, the majority reasoned that modern digital location information differs from the bank records in United States v. Miller and the dialed telephone numbers in Smith v. Maryland. Location History is intensely revealing and is not “shared” with Google in the ordinary sense that a person consciously exposes information to another human being for that person’s independent use. Nor did the Court accept the argument that Location History should be treated differently because users had to take affirmative steps to enable it. In the Court’s view, ordinary use of contemporary smartphones frequently involves affirmative acts, prompts, settings, and permissions, and the Fourth Amendment cannot sensibly turn on an app-by-app, feature-by-feature theory of consent.
At the same time, the decision is not a categorical prohibition on geofence warrants. The Court expressly stopped after deciding that a search occurred. Because the Fourth Circuit had held that there was no search, it had not resolved whether each stage of the warrant was supported by probable cause and sufficiently described with particularity. The Supreme Court therefore remanded those questions. The Court emphasized that a warrant’s reasonableness generally depends on a neutral magistrate’s finding of probable cause and on a scope carefully limited by particularity. In a multi-step digital search, those requirements may have to be examined stage by stage.
The unresolved questions are substantial. Step One required Google to identify every responsive device within the designated area and time, without knowing which device belonged to the robber. Step Two permitted investigators to select a subset and obtain a broader two-hour movement history that extended outside the geofence. Step Three permitted investigators to select a still smaller subset for identifying account information. The warrant said little about the criteria officers would use to decide which users would receive deeper scrutiny. Justice Jackson, joined by Justice Sotomayor, would have resolved at least part of that issue immediately, concluding that Steps Two and Three vested too much discretion in investigators and lacked the judicially determined probable cause and particularity ordinarily required before additional intrusion.
Justice Gorsuch agreed that the government had conducted a search but reached the conclusion through a different route. He continued his project of questioning whether the reasonable-expectation-of-privacy formulation should dominate Fourth Amendment law and argued instead that Chatrie possessed a constitutionally significant property or possessory interest in his Location History. In his view, a person does not necessarily surrender an interest in digital papers or effects merely because a service provider stores them. His concurrence may become important in future cases involving cloud-stored files, personal records, health information, communications, or other digital materials held by intermediaries.
The dissenting opinions warned that the Court was extending Carpenter beyond its stated boundaries. Justice Alito disputed the majority’s treatment of duration and voluntariness and criticized the Court for deciding a major constitutional question in a case involving a Google architecture that had already changed. Justice Barrett separately expressed concern about the majority’s approach to third-party location information. The disagreement reveals that the Court’s digital-privacy jurisprudence remains contested even as a majority has now made clear that precise historical location data receives constitutional protection regardless of whether the government obtains only a short slice of it.
For law-enforcement agencies, the practical lesson is not that reverse-location techniques are forbidden. The lesson is that investigators and prosecutors must approach precise historical location data as Fourth Amendment material from the outset. The affidavit should explain why the defined geographic area and time window are supported by probable cause; why the provider is likely to possess responsive evidence; how uncertainty and location precision will be handled; what objective criteria will govern any narrowing; what additional information may be obtained at each stage; and when renewed judicial approval will be sought. Procedures for uninvolved users, retention, deletion, disclosure, and auditability should be designed before the warrant is executed.
The case also reaches beyond traditional geofence warrants. Its reasoning matters whenever the government uses technology to reconstruct precise movements retrospectively or to search across a large population for persons who match a location-based criterion. That can include some uses of automated license plate reader systems, commercial location-data brokers, application-derived location histories, connected-vehicle data, persistent aerial surveillance, drone systems, real-time crime centers, and AI-enabled data-fusion platforms. The constitutional analysis will remain technology- and fact-specific, but Chatrie strengthens the principle that courts should evaluate what a surveillance system can reveal and how government access changes the scale and character of observation.
Why Chatrie Matters
It transforms the threshold question
Before Chatrie, courts were sharply divided over whether a geofence warrant even implicated the Fourth Amendment rights of an identified user. Some courts treated short-term location history as materially different from the seven or more days of CSLI protected in Carpenter. Others viewed reverse-location acquisition as constitutionally suspect because the government compels a provider to search location records associated with numerous unknown users. After Chatrie, the federal constitutional analysis begins from a clearer premise: once the government acquires an individual’s responsive historical Location History, a search has occurred.
It rejects an easy duration rule
The decision prevents agencies from treating short time windows as categorically outside the Fourth Amendment. Duration still matters to reasonableness and scope, but it is no longer a threshold switch that determines whether the Constitution applies at all. Investigators therefore should not assume that a fifteen-minute, thirty-minute, one-hour, or two-hour request escapes Fourth Amendment analysis merely because it is shorter than the period in Carpenter.
It limits the traditional third-party doctrine
The Court’s reasoning reinforces a broader shift in digital search law. Modern technology routinely places highly personal material in the hands of service providers. The Court again declined to treat third-party possession as a universal waiver of constitutional privacy. The result is especially important for cloud-based policing because many investigative records now exist only because private companies collect, store, process, and retrieve data generated by users.
It focuses attention on the architecture of the warrant
The most consequential future litigation may not concern whether geofence access is a search. It will concern how a reverse-location warrant must be structured. The Court’s remand directs attention to probable cause and particularity “at each step” of the process. That language matters. A constitutionally adequate basis for a limited first-stage anonymous return may not automatically justify expanded movement information or subscriber identification.
It speaks to AI-era surveillance
Modern investigative systems increasingly operate by asking a database to identify the person rather than asking for information about a person already known to police. AI accelerates that model. An agency may know a place, behavior, object, route, face, vehicle characteristic, or time pattern and ask technology to search a population for matches. Chatrie is therefore part of a larger constitutional conversation about reverse searches, data fusion, and mass-query technologies.
Technology Background: Google Location History and Sensorvault
Understanding Chatrie requires understanding the technological architecture that made the investigation possible. At the time of the 2019 robbery, Google offered a feature known as Location History. Users who enabled the feature allowed Google to store a longitudinal record of a device’s locations. Google maintained this information in a repository that came to be commonly described in geofence litigation as “Sensorvault.” The database permitted retrospective queries that could identify devices present within a specified geographic boundary during a specified period.
Location History differed from traditional cellular network records. CSLI generally reflects a device’s interaction with cellular towers and historically has located a phone within a sector or area associated with the serving site. Google’s Location History could integrate multiple location signals, including GPS, Wi-Fi, Bluetooth-related inputs, and cellular information. The Supreme Court’s 2026 opinion emphasized that the resulting records could locate a device within roughly twenty meters and could be recorded approximately every two minutes. The Court also noted evidence that elevation information could sometimes identify a floor within a building.
Those features shaped the constitutional analysis. The location record was not simply a rough indication that a phone had connected to a tower. It could create a fine-grained sequence of movements. In some settings, such data could distinguish between neighboring buildings or between areas within a larger complex. It could reveal arrival, departure, duration, direction of travel, and repeated patterns. The precision and density of the information made Location History more capable than conventional CSLI of exposing the details of physical life.
The database also altered the temporal character of surveillance. Traditional physical observation is ordinarily prospective. Officers decide to follow a particular person or location and devote resources to doing so. A historical location database permits retrospective surveillance. Investigators can decide after an event that a place is important and ask who was there. The target can be discovered by querying the archive rather than selected before surveillance begins.
This retrospective capacity is central to geofence warrants. A conventional location warrant commonly identifies a known person, phone number, account, or device and seeks that target’s location. A geofence warrant reverses the sequence. Officers identify a location and time, compel the provider to find all responsive devices, and then use the returned data to identify one or more people. The warrant thus begins without a named individual whose location history is to be searched.
The Supreme Court also treated the user-facing character of Location History as important. Google users could consult a Timeline of prior movements, edit it, and use it to remember places visited and activities undertaken. The majority therefore analogized the data to cloud-stored emails, documents, photographs, and calendars: records that may reside on a company’s servers but that users reasonably regard as personal materials. That feature strengthened the Court’s conclusion that a user’s privacy interest did not disappear merely because Google physically stored the information.
Accuracy is not certainty
Location estimates are not infallible. A coordinate reported by a provider is not necessarily the literal physical position of the device. Systems generally associate location points with an estimated accuracy radius or confidence measure. A responsive device can therefore appear inside a geofence even when its actual physical position may have been outside the boundary. Conversely, a device physically inside the area may not generate a responsive point during the relevant interval.
For investigators, this is not merely a technical caveat. It affects probable cause, minimization, corroboration, testimony, and disclosure. Maps can create false visual confidence by displaying a precise dot where the underlying system actually reports an estimated range. Affidavits should describe the technology accurately and avoid implying a level of precision unsupported by the provider’s data.
Google’s later architectural change
By the time the Supreme Court decided Chatrie, Google had materially changed how Location History was stored. Google represented that beginning in July 2025 it shifted Location History storage from centralized company servers toward on-device storage, with the consequence that Google could no longer perform the same centralized historical geofence query at issue in the case. That technological development did not moot the constitutional issue because Chatrie’s conviction and suppression claim remained live, and the Court viewed the legal principles as broader than one provider’s current product design.
Facts of the Investigation
On May 20, 2019, a man robbed the Call Federal Credit Union in Midlothian, Virginia. The robber displayed a firearm, demanded money, and left with cash. Witness accounts and security footage provided investigators with information about the offender’s appearance and movements but did not identify him.
Detective David Hylton of the Chesterfield County Police Department reviewed the surveillance evidence. The footage indicated that the robber approached from the direction of a nearby church and that he appeared to possess or use a cell phone during the crime. Investigators pursued conventional leads but did not identify a suspect.
Hylton had previously sought geofence warrants after consulting prosecutors. On June 14, 2019, he applied for another geofence warrant, this time directed to Google. A Virginia magistrate issued the warrant. The geofence was centered on the credit union and had a radius of approximately 150 meters. The initial temporal window covered one hour, beginning thirty minutes before the robbery and ending thirty minutes after.
The location boundary did not contain only the credit union. Like many real-world geofences, it encompassed nearby areas through which innocent people could travel or remain. The record discussed a nearby church and other surrounding locations. That fact became important because a reverse-location warrant necessarily risks collecting information from people whose only connection to the investigation is proximity.
The warrant did not ask Google immediately to identify every user. Instead, it adopted a staged process. Google would first return anonymized identifiers and location points. Investigators would review those data and narrow the set. Google would then provide additional information for a selected subset, and investigators would narrow again before requesting subscriber identities.
Google’s first-stage return produced anonymized data corresponding to nineteen users. Investigators then reduced the group and requested expanded location information for nine users. That expanded return covered two hours and included location points outside the original 150-meter geofence. Officers then selected three users for identification. Google supplied identifying information for those accounts. One belonged to Okello Chatrie.
The records suggested that Chatrie’s device entered the geofenced area shortly before the robbery and traveled toward a residential area afterward. Investigators then conducted additional police work. The geofence result therefore served as an investigative lead rather than the sole evidentiary basis for the prosecution.
A federal grand jury charged Chatrie with robbery and related firearms offenses. He moved to suppress the evidence derived from the geofence warrant, arguing that the process violated the Fourth Amendment. After the district court denied suppression, he eventually entered a conditional guilty plea preserving his right to appeal the geofence issue.
The Three-Step Geofence Process
Identify anonymous devices within the defined area and time
Google searched its Location History repository for devices associated with location points within the 150-meter radius during the one-hour window. The company returned anonymized device identifiers and associated location information for nineteen users. The government did not yet receive names or subscriber details.
Step One creates the central reverse-search problem. The government does not begin by establishing probable cause to search a known person’s location history. Instead, the provider searches a population of users to determine who might have been present. The Supreme Court did not decide whether this first-stage method was sufficiently supported by probable cause or sufficiently particular.
Obtain broader movement history for a selected subset
Investigators selected nine of the nineteen anonymized users for deeper examination. Google returned approximately two hours of location history for those users. The geographic scope expanded beyond the original geofence so that officers could see where the devices came from and where they went.
This stage was more revealing than Step One because it exposed movements outside the crime-scene boundary. That expansion could reveal residences, workplaces, schools, medical facilities, houses of worship, or other sensitive destinations. The constitutional question is whether the warrant itself supplied enough criteria and probable cause for that deeper search, or whether the selection was left too largely to officer discretion.
Reveal subscriber identity for a further subset
After reviewing the expanded location histories, investigators selected three users for identification. Google supplied subscriber information, including data capable of linking the anonymized device identifiers to named account holders. Chatrie was one of the three.
Step Three changed the character of the information from pseudonymous movement data to personally attributable evidence. The Supreme Court did not decide whether the warrant supplied a constitutionally adequate standard for making that transition.
Procedural History
The Call Federal Credit Union in Midlothian, Virginia, was robbed by an unidentified man who appeared to possess a cell phone.
Detective Hylton obtained a warrant from a Virginia magistrate directing Google to carry out a staged search of Location History data associated with the area around the credit union.
The three-stage process moved from nineteen anonymous users to nine expanded histories to three identified accounts, including Chatrie.
Chatrie challenged the geofence evidence under the Fourth Amendment.
The district court denied suppression. It regarded the warrant as constitutionally deficient but concluded that the good-faith exception applied because officers reasonably relied on a warrant issued by a magistrate in an area of law with little precedent.
A divided panel affirmed. The panel majority concluded that Chatrie lacked a reasonable expectation of privacy in the two hours of Location History at issue and therefore no Fourth Amendment search occurred.
The Fourth Circuit vacated the panel disposition and agreed to rehear the case as a full court.
The full Fourth Circuit heard argument over the search question, the validity of the warrant, and good faith.
The court affirmed the district court in a one-sentence per curiam judgment accompanied by extensive separate writings. The judges remained deeply divided over the constitutional merits.
Chatrie sought Supreme Court review.
The Supreme Court agreed to decide the geofence question.
The Court heard argument regarding Carpenter, short-term location information, reverse searching, user voluntariness, and the role of third-party storage.
The Court held that government acquisition of Chatrie’s Location History was a search, vacated the Fourth Circuit judgment, and remanded for analysis of probable cause and particularity at each stage.
Questions Presented and Issues Preserved
The Supreme Court question
The case presented the constitutionality of a geofence warrant used to compel a technology company to identify devices whose stored location history placed them near a crime scene. The threshold dispute was whether the government’s acquisition of Chatrie’s responsive Location History constituted a “search” within the meaning of the Fourth Amendment.
Questions the Court answered
- Does government acquisition of Google Location History implicate a reasonable expectation of privacy? Yes.
- Does the Fourth Amendment apply even though the government obtained only a limited period—approximately two hours—of location history? Yes.
- Does Google’s possession of the information eliminate the user’s Fourth Amendment interest under the third-party doctrine? No.
Questions the Court left open
- Was Step One supported by probable cause?
- Was Step One sufficiently particular, or did it function as an impermissible general search?
- Did the warrant provide constitutionally adequate limits for selecting users at Step Two?
- Was the expanded two-hour, out-of-geofence movement history at Step Two supported by probable cause?
- Did Step Three provide an adequate constitutional basis for subscriber identification?
- Did the warrant improperly delegate probable-cause or particularity judgments to the executing officers?
- Does the good-faith exception ultimately permit use of the evidence even if one or more stages were unconstitutional?
Holding
The Court held that law enforcement officers conducted a Fourth Amendment search when they obtained Chatrie’s Location History from Google. The majority concluded that an individual has a reasonable expectation of privacy in this kind of historical cell-phone location information and that government-compelled access intrudes on that interest.
The holding does not depend on the government obtaining a lengthy period of records. The two-hour period at issue was enough to trigger Fourth Amendment protection. Duration can affect how intrusive a search is and whether the scope is reasonable, but the majority refused to make duration a threshold prerequisite to constitutional coverage.
The holding also does not depend on the information being stored locally on the user’s device. Google’s custody of the records did not eliminate the user’s privacy interest. The Court viewed the data as comparable in important respects to other personal digital material that users store in the cloud while continuing to regard it as private.
Because a search occurred, the Fourth Amendment’s reasonableness requirement applies. In criminal investigations, that generally means a warrant issued by a neutral magistrate upon probable cause and describing the authorized search with particularity, subject to recognized exceptions such as exigent circumstances. The Court remanded so the Fourth Circuit could determine whether this warrant satisfied those requirements at each stage.
The Majority Opinion: Fourth Amendment Analysis
Carpenter as the doctrinal foundation
The majority began with Carpenter v. United States, the 2018 case holding that government acquisition of at least seven days of historical CSLI is a search. Carpenter departed from a mechanical application of the third-party doctrine because cell-phone location records could chronicle a person’s movements with a degree of breadth and retrospective power unknown in earlier cases.
Justice Kagan concluded that everything Carpenter relied upon applied to Location History “as well or better.” Three characteristics mattered: the precision and density of the record, the government’s ability to reconstruct movements retrospectively, and the personal character of the information.
Precision and density
The Court contrasted Location History with conventional CSLI. The record before it indicated that Location History could locate a device within roughly twenty meters, while older CSLI could place a device within much larger sectors. Location History could record a location approximately every two minutes, creating hundreds of potential points in a day. The Court also referenced elevation capabilities that could reveal where a device was located within a multi-story structure.
The doctrinal significance is that precision changes what location data can expose. A rough cell-sector record might show that a phone was somewhere in a neighborhood. More precise data can reveal entry into a specific building or movement between adjacent places. That degree of detail makes it more likely that location records will disclose associations and activities that individuals reasonably expect to remain private from government scrutiny.
Retrospective surveillance
The Court also emphasized that Location History allows law enforcement to reconstruct the past. The government need not decide in advance whom to follow. It can search a historical database after a crime, protest, meeting, or other event and ask which devices were present. That ability reduces the resource constraints that historically limited surveillance.
The majority connected this point to United States v. Jones and Carpenter. In an earlier era, sustained surveillance required teams of officers, vehicles, time, and continuous effort. Digital systems can make retrospective tracking nearly effortless. The reduction in practical friction matters because constitutional expectations developed in a world where many forms of comprehensive surveillance were theoretically possible but practically difficult.
The “personal journal” character of the data
Location History differed from CSLI in another respect: users themselves could use it. The service allowed a user to reconstruct trips, remember restaurants, identify when the user last visited a friend, review vacation travel, or measure distance walked. The Court treated this as evidence that Location History is not merely the provider’s business record. It can serve as the user’s own digital memory.
The majority compared cloud-stored Location History to personal emails, documents, photographs, and calendars. A person does not ordinarily understand the use of cloud storage as an invitation for unrestricted government access. The location record’s placement on Google’s server therefore did not settle the constitutional question.
The Court’s broader technology principle
The majority’s method is functional rather than purely formal. It asks what information the technology creates, how revealing it is, what scale of surveillance it enables, and how users relate to the data. That method is likely to matter in future disputes over location brokers, connected vehicles, wearable devices, cloud records, and AI-generated inferences.
Short-Term Location Data: The Court Rejects a Constitutional Grace Period
One of the government’s principal arguments was that Carpenter protected a lengthy historical record, while Chatrie involved only two hours of location information. Carpenter had expressly reserved whether a shorter period of CSLI might fall outside the Fourth Amendment. The government asked the Court to answer that reserved question by recognizing a short-term safe harbor.
The majority refused. It gave two related reasons.
Even brief location monitoring can expose sensitive facts
First, duration is an imperfect proxy for sensitivity. A single trip can reveal highly private information. A location record can show that a person visited a psychiatrist, reproductive-health provider, criminal-defense lawyer, religious institution, political gathering, treatment center, intimate partner, or other sensitive place. A two-hour window can also capture attendance at a one-time event such as a political rally or gun show.
Geofence technology makes this point more acute because the government can begin with the sensitive place itself. Investigators can draw a perimeter around a location and ask who was present. That is different from merely discovering by happenstance that a known suspect traveled somewhere sensitive.
Fourth Amendment protection is not triggered only by quantity
Second, the Court rejected the premise that an invasion of a protected interest becomes a search only after it passes some quantitative threshold. The majority invoked Kyllo v. United States and other precedents for the proposition that constitutional protection does not depend on the government learning enough intimate detail. If a method intrudes on a protected privacy interest, the fact that the intrusion is brief may affect reasonableness but does not necessarily remove the conduct from the definition of a search.
This distinction is important for agencies. A narrow time window remains a good practice because particularity and proportionality matter. But investigators should use a narrow time window to make a warrant more reasonable, not because they assume short-term data is constitutionally unprotected.
The Third-Party Doctrine After Chatrie
The traditional rule
The third-party doctrine is commonly associated with United States v. Miller and Smith v. Maryland. Those cases held that customers generally lacked Fourth Amendment protection in certain information voluntarily conveyed to banks or telephone companies. The underlying rationale was that a person assumes the risk that information knowingly shared with another may be conveyed to the government.
Carpenter’s limitation
Carpenter declined to extend that doctrine to historical CSLI. The Court emphasized that cell phones are indispensable to modern life and generate location information automatically as they connect to the network. The information is also extraordinarily revealing. The Court therefore treated CSLI as qualitatively different from the records in Miller and Smith.
Location History receives the same protection
In Chatrie, the government argued that Location History should be treated differently because users affirmatively enabled the feature. The Court rejected that distinction. It examined how smartphones operate in practice. Users routinely encounter prompts, settings, and permissions while using apps and services. The fact that some affirmative step is technically required does not necessarily mean the user has meaningfully agreed to unrestricted government access.
The Court also considered what users were told when enabling the feature. The majority noted that Google prompts could encourage users to activate Location History to ensure that services worked correctly without fully conveying the frequency, precision, or potential law-enforcement use of the resulting data. The Court therefore refused to transform a consumer-interface choice into constitutional consent.
What the doctrine still does
Chatrie does not abolish the third-party doctrine. Traditional business records may still fall within it. The case instead reinforces that courts must consider the nature of the information, the circumstances in which it is generated, the degree to which disclosure is genuinely voluntary, and the surveillance capability created by government access.
Separate Opinions
Justice Jackson, joined by Justice Sotomayor
Justice Jackson agreed with the majority’s search holding but would have gone further on the warrant’s validity. Her concurrence focused particularly on Steps Two and Three. In her view, those later stages permitted officers to obtain increasingly sensitive information without requiring a magistrate to make a new probable-cause determination and without providing sufficiently objective criteria to constrain the officers’ choices.
The concurrence treated the staged architecture as constitutionally meaningful. An initial anonymous return may reveal a limited set of devices. The next step exposes a broader movement history. The final step reveals identity. Each escalation can invade privacy more deeply. Justice Jackson argued that a warrant cannot simply hand those escalation decisions to investigators through a vague instruction to narrow the data based on their review.
She also looked at execution history. Investigators initially sought expanded information for all nineteen devices and narrowed the request only because Google pushed back. That fact suggested to her that the provider’s internal policy, rather than the warrant’s judicially imposed limits, provided the most meaningful constraint. The Fourth Amendment ordinarily requires the magistrate—not the provider—to define the authorized search.
For practitioners, the concurrence is a roadmap for future challenges. It suggests that courts may insist on objective stage-specific standards, renewed judicial approval before scope expansion, and a stronger showing of probable cause before identity disclosure.
Justice Gorsuch, concurring in the judgment
Justice Gorsuch agreed that a search occurred but declined to rely on the Katz reasonable-expectation-of-privacy framework. He argued that the Fourth Amendment’s protection of persons, houses, papers, and effects should remain grounded in property and positive-law concepts. In his view, Chatrie retained a sufficient interest in his Location History for government examination of the records to implicate the Fourth Amendment.
His approach treats cloud storage as analogous to entrusting property to another. A bailor does not necessarily lose all rights simply because a bailee possesses the item. By analogy, an individual can retain legal interests in digital records stored by a service provider.
This theory could matter well beyond location information. If adopted more broadly, it may support claims involving cloud documents, health records, digital photographs, communications, or other materials where users possess contractual, statutory, or property-like rights even though a company holds the bytes.
Justice Alito’s dissent
Justice Alito criticized the majority for extending Carpenter to a short period of voluntarily enabled location information. He viewed duration as constitutionally significant and questioned the majority’s treatment of the distinction between automatically generated CSLI and optional Location History.
He also objected to deciding a major constitutional question in the context of technology that Google had already redesigned. In his view, the case risked creating broad doctrine around a historical system that might no longer operate in the same way.
The dissent further emphasized procedural restraint and the good-faith issue. If suppression would not be available because officers reasonably relied on the warrant, Justice Alito questioned the necessity of resolving the broader search question.
Justice Barrett’s dissent
Justice Barrett separately dissented and expressed concern about extending Carpenter to this comparatively short period of third-party location data. Her opinion reflects continuing disagreement within the Court about how far the special treatment of digital location records should move beyond Carpenter’s facts.
What the alignment means
Five Justices joined Justice Kagan’s reasonable-expectation-of-privacy analysis. Justice Gorsuch agreed that a search occurred but used a property-based theory. Thus, six Justices agreed on the judgment that the government’s acquisition implicated the Fourth Amendment, even though they did not all agree on the doctrinal route.
Probable Cause and Particularity: The Questions on Remand
Once the Court determined that the acquisition was a search, the next question became reasonableness. In ordinary criminal investigations, the Fourth Amendment generally requires a warrant issued by a neutral and detached magistrate, supported by probable cause, and describing the scope with particularity.
Probable cause in a reverse-search model
A traditional warrant often identifies the person, house, device, or account to be searched. Probable cause links that known target to evidence of a crime. A geofence warrant works differently. Officers know a place and time but do not know whose device they will search. The provider must examine many users’ records to identify those whose data match the geographic criterion.
That structure creates a difficult question: what exactly is the “place” to be searched for probable-cause purposes? Is it the provider’s database? The subset of records associated with devices within the geofence? Each individual user’s Location History? Courts may characterize the search differently, and the characterization can affect the probable-cause analysis.
The government argued that probable cause existed to believe Google possessed information that would help identify the robber. Chatrie argued that the warrant searched the histories of users for whom there was no individualized reason to suspect criminal involvement. The Supreme Court did not choose between those positions.
Particularity and the general-warrant concern
The particularity requirement exists to prevent wide-ranging exploratory searches. It requires the warrant to confine the government’s authority to the evidence and places justified by probable cause. Geofence warrants test that principle because the initial query can sweep through records of numerous people in order to discover the target.
The key drafting dimensions include geographic radius, time period, data types, accuracy information, the number and nature of devices likely to be captured, the scope of any movement data outside the fence, and the criteria that govern the transition from anonymous information to identity.
Why each stage matters
The Supreme Court repeatedly framed the remaining question in stage-specific terms. Step One, Step Two, and Step Three are not necessarily one indivisible search. The level of intrusion increases. The justification available at the outset may not automatically authorize every later disclosure.
A stronger warrant therefore should articulate why each stage is necessary and what facts permit progression to the next. Where the provider returns a population of anonymous devices, the warrant can identify objective exclusion rules, selection criteria, and a requirement for renewed judicial authorization before expanded data or identities are obtained.
Geographic scope
The radius should be tied to evidence. A map should show the crime scene, relevant entry and exit routes, nearby buildings, public roads, residences, and sensitive locations. The affidavit should explain why the chosen boundary is no larger than necessary to capture the offender’s likely location.
Temporal scope
The time window should be anchored to known offense timing. Investigators should explain whether the requested period covers staging, commission, escape, or accomplice coordination and why each segment is necessary.
Movement outside the geofence
Expanded movement data can be particularly sensitive because it reveals where otherwise anonymous devices traveled before or after the event. The affidavit should explain why this information is necessary and specify limits. A court may view a request for out-of-fence movement differently from a request merely to identify devices present at the scene.
Identity disclosure
The point at which anonymous device data becomes a name, email address, phone number, or account identifier should receive independent attention. If investigative analysis can eliminate innocent devices while they remain anonymous, delayed identification reduces privacy intrusion. A staged warrant can require additional judicial approval before the provider reveals identities.
The Good-Faith Exception
The district court denied suppression under United States v. Leon. Even when a warrant is later found invalid, evidence may remain admissible when officers acted in objectively reasonable reliance on a warrant issued by a neutral magistrate, unless one of the recognized circumstances defeating good faith is present.
The district court emphasized the novelty of geofence warrants in 2019. Detective Hylton had previously sought similar warrants, consulted prosecutors, and obtained judicial approval. At the time, courts had not developed a settled body of geofence precedent. The district court therefore concluded that exclusion would have little deterrent value.
The Fourth Circuit’s en banc opinions repeatedly returned to good faith. Chief Judge Diaz would have affirmed on that narrow ground without resolving the constitutional merits. Other judges agreed that good faith applied even while disagreeing about whether the warrant constituted a search or violated the Fourth Amendment.
The Supreme Court did not finally decide the good-faith question. Its decision may, however, change the analysis for future warrants. Once the Supreme Court has declared that government acquisition of precise historical Location History is a search, officers can no longer rely on the proposition that the Fourth Amendment is inapplicable. Future warrants also will be evaluated against a developing body of case law concerning reverse searches, stage-specific probable cause, and particularity.
That distinction is important operationally. Good faith is an exclusionary-rule doctrine, not an authorization to ignore constitutional uncertainty. Agencies should draft to satisfy the Fourth Amendment rather than assume evidence will be saved later by Leon. The more clearly precedent identifies a defect, the harder it may become to characterize reliance on the same defect as objectively reasonable.
How Chatrie Fits with the Supreme Court’s Technology Cases
| Case | Technology / Record | Core Principle | Connection to Chatrie |
|---|---|---|---|
| Katz v. United States | Electronic listening | Fourth Amendment protects people and reasonable privacy, not only physical places. | Provides the expectation-of-privacy framework used by the majority. |
| United States v. Knotts | Beeper tracking on public roads | Short-term augmentation of visual tracking did not constitute a search on the facts presented. | Government relied on Knotts; Chatrie distinguished database-driven retrospective tracking. |
| United States v. Karo | Beeper tracking into a private residence | Technology revealing information from inside protected space can implicate the Fourth Amendment. | Supports attention to place and the sensitivity of technologically revealed location. |
| Kyllo v. United States | Thermal imaging | Government cannot avoid the Fourth Amendment merely because technology reveals limited detail. | Chatrie uses Kyllo to reject a quantity-of-information threshold. |
| United States v. Jones | GPS tracking | Physical installation of GPS was a search; concurrences highlighted prolonged location monitoring. | Supplies the intellectual foundation for concern about aggregated movements and short-term sensitive trips. |
| Riley v. California | Cell-phone contents | Digital devices differ in scale and sensitivity from physical objects. | Chatrie draws on Riley’s recognition of the pervasive role of smartphones. |
| Carpenter v. United States | Historical CSLI | Government acquisition of at least seven days of CSLI is a search despite third-party possession. | Direct doctrinal foundation; Chatrie extends protection to more precise Location History and rejects a short-term safe harbor. |
From tracking a known suspect to discovering the suspect
Jones and Carpenter largely involved government efforts to track known people. Chatrie addresses a different model: the government searches location information to discover who the suspect might be. That reverse-search feature increases the relevance of general-warrant history because uninvolved people become part of the initial query.
From scarcity to scale
The Court’s technology cases increasingly recognize that digital systems change not only the accuracy of surveillance but its cost and scale. A technique that once required many officers can be performed against a database in seconds. Constitutional analysis therefore cannot always rely on the assumption that practical resource limitations will constrain government access.
Practical Guidance for Law Enforcement
Start with a warrant mindset
After Chatrie, precise historical location data should be treated as protected unless a recognized exception applies. Investigators should involve prosecutors or agency counsel early, particularly when using novel reverse-search techniques.
Define the investigative theory
The affidavit should explain why the offender likely possessed a device that would generate responsive data. Surveillance video, witness statements, communications evidence, or the nature of the offense may support that inference. Avoid generic assumptions that every person necessarily carries a traceable phone.
Justify the geography
Explain why the selected radius captures the relevant crime scene and likely path without unnecessarily including adjacent homes, businesses, schools, churches, medical facilities, or unrelated public spaces. If sensitive locations are necessarily included, identify them and explain why the boundary cannot reasonably be drawn more narrowly.
Justify the time
Use known timestamps. Distinguish between time necessary to capture approach, commission, and escape. Avoid adding large buffers merely for convenience.
Describe the technology accurately
Explain the provider’s data source, precision, uncertainty, retention, and query process. Do not describe an estimated coordinate as an exact physical location. Include accuracy radius information if available.
Limit the initial return
Request the minimum fields needed to assess whether a device could be associated with the offense. Where possible, begin with anonymized identifiers.
Use objective narrowing criteria
State in advance how officers will exclude or retain devices. Criteria can include temporal consistency, path of travel, stationary presence at a business, repeated points indicating an uninvolved resident or employee, or movement inconsistent with the suspect’s known conduct. Criteria should be tied to evidence rather than intuition.
Consider renewed judicial approval
A particularly strong structure is to require investigators to return to the issuing judge before receiving expanded movement history or subscriber identity. This places the magistrate—not the provider or investigator—at the center of escalation decisions.
Protect uninvolved users
Plan for deletion, segregation, retention limits, access controls, and dissemination restrictions. Document when a device is eliminated and what happens to its data.
Preserve the execution record
Keep provider correspondence, returned datasets, analyst notes, selection decisions, maps, accuracy information, and the exact basis for each narrowing step. Later litigation may focus as much on execution as on the face of the affidavit.
Corroborate before enforcement action
A location match is a lead, not identity proof. A phone may be carried by someone other than the subscriber, left in a vehicle, lent to another person, or inaccurately located. Investigators should corroborate device ownership, possession, movement, opportunity, and other evidence before arrest or search.
Geofence / Reverse-Location Warrant Drafting Checklist
- Identify the offense precisely. State the crime, date, location, and known sequence of events.
- Explain why location data will contain evidence. Describe facts supporting the inference that the offender possessed a device or used a location-enabled service.
- Identify the provider and data source. Explain what data is believed to exist and why the provider possesses it.
- Describe the technology. Include collection method, frequency, precision, uncertainty, and retention limitations.
- Define the geographic boundary. Use maps or coordinates and explain why the radius is tied to the offense.
- Account for nearby sensitive locations. Identify homes, schools, churches, hospitals, clinics, political offices, or other sensitive places within or near the fence.
- Define the temporal window. Tie each minute or buffer to staging, offense conduct, escape, or another articulable need.
- Specify Step One data fields. Request only what is necessary to identify responsive anonymous devices.
- State accuracy handling. Require production of available confidence or accuracy information and explain how officers will interpret it.
- Define objective exclusion criteria. Identify when devices will be removed from further scrutiny.
- Define objective inclusion criteria. Explain what facts permit a device to progress to the next stage.
- Limit out-of-geofence data. If expanded movement is necessary, specify duration and purpose.
- Consider judicial reauthorization. Require a return to the court before Step Two or Step Three when feasible.
- Delay identifying information. Keep users anonymous until identity is necessary and constitutionally justified.
- Define retention and deletion. State what will happen to information concerning eliminated users.
- Limit dissemination. Restrict unrelated user information from entering intelligence systems or investigative reports.
- Document provider modifications. If the provider narrows or alters the process, preserve the correspondence and seek judicial clarification where material.
- Preserve audit records. Maintain a reproducible record of each query, return, filtering decision, and disclosure.
- Address exigency separately. If emergency facts may justify faster access, articulate them rather than relying on the ordinary warrant process.
- Review current precedent. The law remains rapidly developing and may vary by jurisdiction and technology.
Prosecutor and Agency-Counsel Review Checklist
Before presenting a reverse-location application, reviewing counsel should test the warrant from the perspective of a suppression judge rather than merely asking whether the technique is useful.
- Does the affidavit establish a fair probability that the provider’s location records will contain evidence related to the crime?
- How many unrelated users are reasonably expected to fall within the initial search?
- Can the radius be reduced without sacrificing the investigative objective?
- Can the time window be reduced?
- Is the provider’s technical process described accurately and currently?
- Does the application distinguish location estimate from exact physical presence?
- What facts permit officers to move a device from Step One to Step Two?
- What facts permit identification at Step Three?
- Who makes each narrowing decision: the provider, officers, or the court?
- Does the judge retain meaningful control over escalating intrusion?
- Are movements outside the geofence necessary? If so, why?
- What happens to data belonging to people eliminated from suspicion?
- Does the warrant avoid collecting protected First Amendment association unnecessarily?
- Are state constitutional or statutory protections more restrictive than federal law?
- Would an ordinary targeted investigative technique accomplish the objective with less intrusion?
Litigation Issues for Defense Counsel and Courts
A complete analysis of Chatrie should also identify the principal suppression issues likely to be raised against future warrants. Doing so helps agencies draft better warrants and helps courts identify the actual points of constitutional dispute.
Standing and personal rights
Fourth Amendment rights are personal. A defendant generally must establish that government conduct invaded the defendant’s own protected interest. After Chatrie, a person whose Location History was acquired has a strong basis for arguing that a search of that person occurred. Questions may remain for users whose records were queried internally by a provider but never disclosed to the government.
General-warrant theory
Defense counsel may argue that a reverse search resembles a general warrant because it begins by examining the records of numerous unknown people without individualized suspicion. The government will respond that the warrant is limited by a specific place, time, crime, and category of evidence. The remand in Chatrie leaves that debate open.
Nexus and probable cause
A defense challenge may ask whether the affidavit established more than a possibility that the offender possessed a Google account with Location History enabled. If the warrant captures all responsive users, counsel may argue that probable cause to believe the perpetrator’s data is present does not equal probable cause to search every innocent person whose data is also present.
Particularity
Challenges may focus on the size of the geofence, the duration, the breadth of expanded movement, the absence of objective narrowing rules, and officer discretion over identification.
Execution beyond the warrant
Even a facially valid warrant can be executed unreasonably. Litigation may examine whether investigators requested more data than authorized, changed narrowing criteria, retained unrelated information, or used returned data for purposes outside the warrant.
Good faith
Future defendants will argue that Chatrie and subsequent cases put officers on notice of constitutional limits, weakening good-faith arguments for warrants that repeat known defects. Prosecutors will emphasize continuing uncertainty on the stage-specific questions not decided by the Supreme Court.
AI, Reverse Searching, and Data Fusion
The deepest significance of Chatrie may emerge as police technology moves from single-database searching to AI-assisted fusion. A geofence query asks a database to identify devices matching a place-and-time condition. AI systems can perform far more complex versions of the same logic: identify people whose movements, vehicles, communications, purchases, images, or behavioral features match an investigative pattern.
Reverse searching becomes easier
Historically, investigators often began with a known suspect and asked what information existed about that person. Modern analytics can begin with an event and search across populations to discover potential suspects. Examples include reverse keyword searches, reverse location searches, unknown-face facial recognition, ALPR pattern searches, gait matching, and cross-database identity resolution.
AI does not eliminate the need for constitutional articulation
An algorithmic recommendation does not itself establish probable cause. Agencies should understand what the system queried, what population was searched, what matching threshold was used, and what false-positive rate or uncertainty is associated with the output.
Data fusion may magnify privacy
Combining several individually limited datasets can create a highly revealing profile. A short location trace may become far more sensitive when merged with ALPR sightings, social-media posts, camera footage, dispatch data, commercial broker records, or employment information. Courts may increasingly examine the surveillance system as a functional whole rather than treating each source in isolation.
Explainable narrowing
If AI is used to rank or eliminate responsive devices, the decision process must be auditable. Investigators should preserve the model or rule set, inputs, thresholds, timestamps, and analyst interventions. A warrant should not delegate constitutionally significant narrowing to an opaque vendor algorithm without understanding how that process works.
Human validation
AI output should generate leads, not conclusive identity. Device-to-person attribution, location accuracy, shared accounts, borrowed phones, spoofed data, and model error all require independent corroboration.
Implications for ALPR, Pole Cameras, Drones, and Persistent Surveillance
Automated license plate readers
ALPR systems typically collect observations of vehicles in public rather than cell-phone location history. Chatrie does not hold that every ALPR query is a search. Its reasoning nevertheless matters when agencies use very large historical datasets to reconstruct movements, identify all vehicles present at a sensitive location, or conduct reverse searches against people not previously suspected.
The more an ALPR network permits retrospective reconstruction of an individual’s movements across time and geography, the closer the functional privacy concern moves toward Jones, Carpenter, and Chatrie. Agencies should distinguish ordinary plate lookup from pattern-of-life analytics and mass reverse queries.
Pole cameras
Long-term pole-camera cases remain divided. Cameras generally observe what is exposed to public view, but continuous recording and retrospective review can create an aggregated record of activity around a home or other sensitive place. Chatrie reinforces the proposition that digital aggregation and retrospective capacity can alter the constitutional significance of surveillance, even when individual observations appear mundane.
Drones and persistent aerial surveillance
A drone flight of short duration over a public area is not the same technology as Google Location History. But persistent aerial systems capable of recording an entire city and reconstructing movements after the fact raise similar scale questions. Leaders of a Beautiful Struggle v. Baltimore Police Department illustrates that concern in the Fourth Circuit.
Real-Time Crime Centers
RTCC platforms increasingly integrate cameras, ALPR, gunshot detection, mapping, CAD, criminal-intelligence databases, and commercial datasets. A single query may trigger searches across multiple sources. Agencies should document the legal authority and retention rules for each source and be alert to situations where the fused result reconstructs movement or association in a manner more revealing than any individual input.
Agency Policy Recommendations
Define reverse-location technology
Policy should distinguish targeted location requests from reverse-location searches that begin with a place, time, or behavioral condition and search a population to identify unknown persons.
Require legal review
Reverse-location applications should receive prosecutor or agency-counsel review before submission except in genuine emergencies. The review should address federal and state constitutional law, statutes, provider requirements, and current appellate precedent.
Require written scope justification
The investigative file should document why the geographic and temporal boundaries were selected and why less intrusive alternatives are inadequate or impractical.
Mandate minimization
Policies should require prompt segregation or deletion of data concerning users eliminated from the investigation, absent an independent lawful basis for retention.
Protect First Amendment activity
Reverse-location searches near political demonstrations, houses of worship, news organizations, clinics, advocacy offices, or other sensitive locations should receive heightened review because location can reveal protected association and expression.
Audit use
Agencies should maintain a log of reverse-location applications, approvals, providers, scope, number of responsive users, number of identified users, outcomes, and any complaints or suppression rulings. Periodic audit can identify unnecessarily broad practices.
Vendor transparency
Contracts with analytics providers should require disclosure of data sources, collection methods, accuracy limitations, retention, query logging, and whether the vendor uses agency queries or results to improve commercial models.
Training
Investigators should understand that a device location is not synonymous with a person’s physical presence. Training should address accuracy radii, device sharing, stale data, identity attribution, false positives, corroboration, and disclosure obligations.
Instructor and Training Notes
Teaching objective 1: Separate “search” from “reasonableness”
Students often collapse these concepts. Chatrie is a strong teaching case because the Supreme Court clearly held that a search occurred while expressly leaving open whether the search was reasonable. A search is not automatically unconstitutional; a warrant is not automatically valid.
Teaching objective 2: Compare known-target and unknown-target surveillance
Ask students to compare a warrant for seven days of a known suspect’s CSLI with a one-hour geofence that searches nineteen unknown users. Which is more intrusive? The answer depends on what dimension of privacy is being measured: duration, precision, population size, identity, or retrospective power.
Teaching objective 3: Identify the constitutional decision-maker
Focus on who decides to escalate from Step One to Step Two and from Step Two to Step Three. If the warrant leaves the decision entirely to officers or a provider, what role remains for the magistrate? Justice Jackson’s concurrence makes this question concrete.
Teaching objective 4: Distinguish provider policy from constitutional limits
Google’s internal narrowing practices sometimes imposed restraints beyond the face of the warrant. Students should understand that private provider policy can reduce actual intrusion but does not necessarily cure a constitutionally deficient warrant.
Teaching objective 5: Apply the rule to new technology
Present a hypothetical in which an agency has access to ten years of ALPR data and asks the system to identify every vehicle that visited a mosque, gun show, and particular residence during a two-week period. Ask whether Chatrie, Carpenter, or Jones changes the analysis even though each license plate was observed in public.
Shield Practice Notes
- Treat precise historical cell-phone location data as Fourth Amendment material.
- Do not rely on a short time window as a search exception. Narrow time helps particularity; it does not eliminate constitutional protection.
- Do not assume cloud storage defeats privacy. Third-party possession is not dispositive for highly revealing digital records.
- Build probable cause around the actual reverse-search process. Explain why the provider is likely to possess evidence and why the selected population and scope are justified.
- Draft stage by stage. Step One, expanded movement, and identification should each have an articulated legal basis.
- Use objective narrowing criteria. Avoid leaving escalation to undefined officer discretion.
- Consider returning to the judge before expanded data or identity disclosure.
- Account for sensitive locations. Churches, clinics, political events, schools, residences, and news organizations deserve particular attention.
- Preserve location uncertainty. Never convert estimated coordinates into false precision.
- Distinguish a device from a person. Corroborate possession and identity.
- Minimize data about uninvolved people. Retention should require a lawful reason, not mere technical availability.
- Document the provider’s role. If a provider imposes narrowing that the warrant does not require, preserve that history and consider whether judicial clarification is necessary.
- Do not depend on good faith as a drafting strategy. Good faith addresses suppression after a mistake; it is not a substitute for constitutional design.
- Apply Chatrie’s functional reasoning to AI and data fusion. Ask what the system can reconstruct, whom it searches, and how easily it can retrospectively expose private patterns.