Geofence Warrants and Location Sweeps in Criminal Cases
A geofence warrant works backward. Instead of seeking location records for a known suspect, it requires a technology company to search its location…
Reverse-location searches, device sweeps, anonymized identifiers, staged disclosure, probable cause, particularity, accuracy radii, homes and sensitive places, third-party records, good faith, and suppression
Geofence Warrants and Location Sweeps in Criminal Cases
A geofence warrant works backward. Instead of seeking location records for a known suspect, it requires a technology company to search its location database for devices near a place during a specified time and then helps investigators identify selected users. In 2026, the United States Supreme Court held that obtaining a user’s location information through this process is a Fourth Amendment search. Whether a particular multi-step warrant satisfies probable cause and particularity must be tested at every stage.
Chatrie establishes that geofence acquisition is a Fourth Amendment search
In Chatrie v. United States, 609 U.S. ___, slip op. at 2, 10, 29–32 (2026), the Supreme Court held that an individual has a reasonable expectation of privacy in records about the location of the person’s cellphone. Police intrude on that protected interest when they demand the information from a technology company, even for a limited time.
Provider possession does not eliminate privacy
Chatrie applies the digital-age reasoning of Carpenter v. United States, 585 U.S. 296, 309–22 (2018). A person does not surrender all Fourth Amendment protection merely because a provider collects location data as part of a service. The sensitivity, comprehensiveness, automatic generation, and retrospective reach distinguish these records from ordinary business information.
The Court did not decide whether Chatrie’s warrant was reasonable
The Supreme Court left the Fourth Circuit to decide whether the warrant satisfied particularity and probable cause at every stage. It also did not decide whether the good-faith exception ultimately permits admission. A current guide should not misstate Chatrie as banning geofence warrants or approving the three-stage process.
The decision applies beyond lengthy tracking
Chatrie rejects the argument that a short acquisition necessarily avoids the Fourth Amendment. The warrant sought an hour of data within the initial perimeter and two hours for selected anonymous users. Even that limited duration constituted a search because the government accessed protected cellphone-location information.
Geofence capability has changed, but the legal issues remain
The Chatrie opinion notes that Google changed Location History storage in July 2025 from its central servers to individual users’ devices and represented that it can no longer answer the same kind of Location History geofence warrant. Other providers and data sets may still support reverse-location demands, and legacy investigations remain in court. The constitutional rule is about government acquisition of protected location data, not one product name.
North Dakota procedure still governs a state warrant
N.D.R.Crim.P. 41 requires probable cause and particular description and allows warrants for electronically stored information. A state magistrate must independently authorize the search; a provider’s ability to generate a return does not establish constitutional validity.
A geofence search identifies suspects from a population of unknown users
Traditional warrants begin with a person, account, or place
Investigators ordinarily develop probable cause concerning a known target and request that person’s records. A geofence demand begins with a place and time, searches provider data for everyone fitting those parameters, and uses resulting movements to decide whose identity to reveal. That reverse structure creates the risk that innocent people become investigative subjects because they were nearby.
The historical Google protocol used three stages
As described in Chatrie, stage one returned anonymous device identifiers with latitude and longitude, timestamps, estimated accuracy, and source for devices within the perimeter and time. At stage two, investigators selected anonymous devices and received a longer interval of movements, including points outside the fence. At stage three, the provider disclosed identifying account information for a final subset.
“Anonymized” data can still reveal identity
Movement from a residence to a workplace, school, clinic, place of worship, or repeated destination can identify a person before subscriber information is formally disclosed. An anonymous device ID is also a stable handle that permits sorting and correlation. The constitutional intrusion does not begin only when a name appears.
The provider searches its entire eligible database
The government may receive only several records, but the provider must compare the warrant’s parameters against a much larger collection to determine which users match. That database-wide operation is central to breadth. The affidavit should explain eligible users, opt-in or default settings, retention, data sources, and what provider process will actually occur.
Stage two can be more revealing than the original fence
Extended-location data may show where a person came from and went afterward, including private homes and sensitive places outside the authorized perimeter. In Chatrie, the second-stage production was not geographically bounded by the fence. A warrant should not let investigators enlarge geography and duration merely by selecting anonymous IDs under undefined standards.
Other reverse-location requests may use different architecture
Cellular tower dumps, advertising-identifier data, vehicle telematics, app location, ride records, Bluetooth detections, Wi-Fi logs, license-plate systems, and provider “nearby device” searches are not technically identical. Label the source correctly. Chatrie’s core search holding applies to protected phone-location data, while probable cause, particularity, accuracy, and statutory process depend on the system.
Probable cause must justify searching the people inside the fence
Probable cause that a crime occurred is not enough
A crime scene establishes that an offense happened at a place, but a geofence warrant searches location records belonging to multiple users. The affidavit should establish why the perpetrator likely carried a trackable device, why the provider likely holds responsive data, and why the selected area and time will distinguish involved devices from innocent traffic.
Evidence that the offender used a phone is important
In Chatrie, surveillance indicated that the robber appeared to use a phone while approaching the credit union and might have communicated with an accomplice. Comparable facts can strengthen the device nexus. A generic assertion that almost everyone carries a phone is less particularized and risks treating ordinary technology use as probable cause to search everyone nearby.
Density and ordinary activity change the probability
A narrow rural scene at an unusual hour differs from a downtown block containing apartments, offices, restaurants, transit, and public sidewalks. Estimate residents, employees, customers, worshippers, students, patients, travelers, and traffic. The expected innocent population bears directly on whether presence meaningfully indicates involvement.
Time should match the offense opportunity
The interval should be grounded in surveillance, dispatch, access logs, witnesses, transaction timestamps, travel time, and the crime’s duration. Padding before and after the event adds innocent users and movements. If investigators seek a longer interval to trace approach and departure, the affidavit should justify the expansion and the magistrate should authorize it expressly.
Presence near a crime does not create individualized probable cause
Ybarra v. Illinois, 444 U.S. 85, 91–96 (1979), holds that a warrant to search a tavern and bartender did not authorize searches of every patron without individualized probable cause. A geofence involves records rather than pat-downs, but Ybarra illustrates the constitutional problem with converting proximity to a place into suspicion of every person present.
Each stage needs support
Chatrie directs attention to probable cause at each step. The government should justify the initial device sweep, any longer and geographically broader movement history, and the ultimate identification demand. Earlier inclusion in a broad anonymous set does not itself create probable cause for expanded tracking.
North Dakota’s nexus rule applies
State v. Leavitt, 2015 ND 146, ¶¶ 5–9, requires a nexus between the place searched and evidence sought. Circumstantial evidence may establish it, but the magistrate needs case-specific facts. For a reverse-location search, the affidavit should connect the offender, a carried device, the data source, area, time, and requested disclosure.
Particularity must constrain geography, time, data, stages, and discretion
The fence must be drawn from facts, not convenience
Identify the center coordinate, shape, radius or polygon, map scale, building boundaries, roads, parking, vertical dimension, and accuracy expansion. A circular fence around a business may sweep adjacent residences, another floor, a church, clinic, or public route. Include maps showing the real built environment.
The warrant should fix objective narrowing criteria
Instructions that officers will “attempt to narrow” a list leave essential decisions to the search team. Criteria might address minimum dwell time, entry and exit matching surveillance, exclusion of known employees or residents, direction of travel, repeated points, or other facts. The magistrate—not officers after seeing private movements—should approve the rules controlling expanded disclosure.
Judicial approval between stages can reduce discretion
A staged warrant can require officers to return to the issuing court with the anonymous stage-one results and establish cause before receiving broader movements or identities. That procedure is not automatically sufficient, but it places the neutral magistrate between increasingly intrusive steps instead of delegating the progression entirely to investigators.
Accuracy radii can extend beyond the drawn perimeter
If a reported point has a 100-meter accuracy radius, the device’s true location may lie anywhere within that circle. Providers may include a device because its estimated location or uncertainty overlaps the fence, even when the user was outside. The warrant should state how uncertainty affects inclusion, and officers should not portray a dot as exact presence.
Sensitive places require special scrutiny
Homes, medical facilities, houses of worship, schools, political meetings, shelters, legal offices, and treatment programs can reveal intimate associations. Their inclusion increases both innocent population and privacy harm. The affidavit should identify known sensitive sites and explain why the search cannot exclude or minimize them.
The return should not exceed the categories authorized
Location points, timestamps, source, accuracy, device identifiers, subscriber names, email addresses, phone numbers, account creation, and payment are separate categories. A warrant should say what each stage may disclose. The provider’s standard return does not expand the signed judicial command.
General warrants are the historical concern
Stanford v. Texas, 379 U.S. 476, 481–86 (1965), emphasizes the Fourth Amendment’s hostility to general warrants and exploratory rummaging. Reverse searches deserve careful limits because investigators begin without knowing whose private records will be searched.
The signed warrant must contain or incorporate the limits
Groh v. Ramirez, 540 U.S. 551, 557–63 (2004), rejects reliance on detail in an application that the facially deficient warrant did not incorporate. Maps, stage descriptions, criteria, dates, data fields, and identification rules should be part of the operative authorization delivered to the provider and officers.
Location-history points are estimates, not GPS truth
Several signals can contribute to a point
A provider may estimate location from GPS, Wi-Fi access points, Bluetooth, cellular networks, IP information, device sensors, or a blend. The source can change from point to point. Obtain the provider’s field definitions and system documentation applicable when the data was collected.
The accuracy value is part of the evidence
A latitude-longitude coordinate without its uncertainty creates false precision. Plot the full accuracy radius and determine whether it crosses the fence, buildings, roads, floors, or alternative locations. Ask whether the reported value represents a confidence estimate and what probability the provider assigns.
Sampling intervals leave gaps
Location History in Chatrie recorded data approximately every two minutes, not continuously. A device may enter and leave between points. Power state, permissions, background limits, network access, battery settings, and user configuration affect reporting. Absence of a point is not necessarily absence of a device.
The phone’s location is not automatically the person’s location
A device can be loaned, stolen, left in a vehicle, carried by another person, or associated with a shared account. Conversely, a person may carry multiple devices. Surveillance, witnesses, possession, account sessions, device identifiers, and later corroboration must connect device and person.
Maps can hide uncertainty
Investigative graphics may display a small pin at the center coordinate and omit the accuracy circle. Zoom, projection, building footprints, and basemap error can make a point appear inside a room. Defense exhibits should show scale, uncertainty, source, time, and alternative locations fairly.
Provider changes affect reproducibility
Algorithms, retention, user controls, storage architecture, and export formats change. A current support page may not describe a 2019 system. Preserve historical documentation, warrant-response guides, declarations, and source fields. Google’s 2025 move to on-device Location History is a clear example of architecture changing over time.
Execution must follow the warrant’s staged rules
Reconstruct what the provider actually searched
Request the warrant, provider submission, portal receipt, correspondence, maps or coordinates in machine-readable form, time zones, stage-one return, expansion request, stage-two return, identification request, final response, cover letters, and field guides. A detective summary cannot substitute for the production chain.
Identify who narrowed the list and how
Obtain spreadsheets, scripts, maps, notes, emails, meetings, criteria, exclusion decisions, and audit logs. If several investigators reviewed anonymous tracks, document each person’s choices. Compare actual criteria with those the magistrate approved.
Expansion outside the fence requires authority
Stage two may expose travel to homes, schools, hospitals, workplaces, and other locations. Determine whether the signed warrant specified the expanded duration and lack of geographic limit. If officers selected devices based on information not presented to the judge, identify the resulting search branch.
Identity disclosure is another search step
Linking a track to a name, email, phone, or account allows investigators to combine the location record with databases and personal history. Chatrie directs courts to evaluate each stage. Anonymous inclusion alone should not automatically authorize identification under criteria invented after the return.
Later warrants may depend on the sweep
Geofence results often lead to subscriber checks, surveillance, interviews, phone warrants, home warrants, arrest, and statements. Create a derivative-evidence tree. If a stage was unlawful, test independent source, attenuation, inevitable discovery, and whether later affidavits accurately described uncertainty.
Provider overproduction does not expand judicial authority
If the company returns a larger area, longer interval, more fields, or more users, investigators should segregate excess material and seek guidance before review. Government cannot convert a provider error into an open-ended search.
Related doctrines do not erase the stage-by-stage warrant inquiry
Investigative necessity is relevant but not probable cause by itself
Geofence applications often explain that witness interviews, surveillance footage, tips, and database checks failed to identify a suspect. Exhaustion can explain why officers seek an unusual tool and may bear on reasonableness. It does not create probable cause to search every nearby device. The affidavit still needs facts connecting the offender, a device, provider data, geography, and time.
Seriousness of the offense does not enlarge the warrant power
Armed robbery, homicide, kidnapping, and immediate threats understandably generate intensive investigation. The Fourth Amendment’s probable-cause and particularity requirements continue to apply. Offense severity may support emergency action when facts show imminent danger, but it cannot substitute for individualized judicial analysis once officers seek historical records.
Exigent circumstances require a genuine emergency
Chatrie notes that established exigency doctrine remains available when law-enforcement needs are sufficiently compelling. A missing child, active shooter, imminent attack, or urgent threat may support prompt acquisition tailored to the emergency. The State should establish immediacy, why ordinary process was inadequate, the smallest practical area and interval, and when continued review ceased or shifted to warrant authority.
Consent by one user does not authorize a population sweep
A known victim or witness may voluntarily share personal location history. That consent permits access within its scope; it does not authorize provider disclosure about every unknown device in the same area. Investigators can use consensual data to refine time, movement, and geography, but a broader reverse search needs an independent legal basis.
Investigators should preserve elimination evidence
Officers may exclude devices belonging to employees, residents, delivery drivers, first responders, or witnesses. Preserve how those users were identified, whether their tracks contradicted assumptions, and why other users remained. Selective preservation of only the eventual suspect’s track can conceal false positives and prevent testing whether narrowing criteria were applied consistently.
A later confession does not retroactively validate the search
Geofence identification may lead to interrogation and admission. The confession can be powerful evidence, but the warrant’s validity is judged from the information presented before execution. If the location search was unlawful, courts must separately analyze whether the statement and later evidence were fruits, whether warnings and voluntariness were sufficient, and whether attenuation or an independent source breaks the causal chain.
Standing language should reflect Chatrie’s protected-interest holding
After Chatrie, an identified user has a protected privacy interest in the acquired cellphone-location records. Litigation should focus on whose data government obtained and how it was used, rather than assuming provider possession defeats the claim. If a defendant challenges a broader sweep of other users, remedy and third-party-rights questions remain distinct from whether acquisition of the defendant’s own track was a search.
Litigating a geofence warrant in North Dakota
Obtain the complete sealed record
Seek the application, affidavits, sworn testimony, maps, attachments, signed warrant, preservation request, provider communications, returns, staged requests, narrowing work, supplemental warrants, returns and inventories, nondisclosure materials, and later derivative warrants. Move for appropriate access to sealed materials needed for the defense.
Build a user-and-stage matrix
For each stage, record the geographic area, interval, accuracy treatment, data fields, eligible database, users returned, narrowing rule, person making the choice, judicial review, expanded movements, identities disclosed, and evidence generated. This shows whether cause and particularity weakened as intrusion grew.
Consider a Franks challenge
Under Franks v. Delaware, 438 U.S. 154, 155–56, 171–72 (1978), a substantial preliminary showing of a material intentional or reckless affidavit falsehood can require a hearing. Geofence issues can include inaccurate precision, omitted homes or sensitive places, understated user counts, misdescribed opt-in, false claims of phone use, omitted alternative suspects, or misleading narrowing descriptions.
Good faith remains a separate question
United States v. Leon, 468 U.S. 897, 914–23 (1984), supplies a federal good-faith exception with limits. Chatrie expressly left good faith for further consideration. Timing, available precedent, provider guidance, officer consultation, affidavit candor, facial breadth, judicial role, and execution beyond the warrant all matter.
Novelty does not permanently excuse defective warrants
An officer’s reliance can be more plausible when a tool is genuinely new and courts have given little guidance. After Chatrie, officers know acquisition is a search and that probable cause and particularity apply at every stage. Continued use of undefined narrowing and broad expansions becomes harder to characterize as reasonable uncertainty.
Request technical expertise and source data
An expert may evaluate coordinate systems, time zones, accuracy radii, sampling, provider fields, mapping, overlap calculations, scripts, user counts, and device attribution. Preserve native structured data and processing code. A screenshot of dots is not an adequate forensic production.
File under Rule 12 and preserve specific remedies
N.D.R.Crim.P. 12 generally requires pretrial suppression issues when reasonably available. Identify the invalid stage, affected users and records, later fruits, Franks issues, execution excess, and good-faith response. Ask for findings at each stage rather than a single conclusion about the warrant as a whole.
Frequently asked questions about geofence warrants
What is a geofence warrant?
It is a reverse-location demand asking a provider to identify devices within a defined area and time, often followed by broader movement data and subscriber identification for selected anonymous users.
Did the Supreme Court ban geofence warrants?
No. Chatrie holds that acquiring the location data is a Fourth Amendment search. It left lower courts to decide whether the particular multi-stage warrant satisfied probable cause and particularity.
Does a short time interval avoid the Fourth Amendment?
No categorical short-duration exception applies to this technique. Chatrie treated acquisition of the limited Location History data there as a search.
Does sharing location with a provider eliminate privacy?
No. Chatrie and Carpenter reject automatic application of the third-party doctrine to sensitive cellphone-location records.
Are anonymous device records private?
They can be. Stable anonymous tracks reveal movements and may identify homes, work, schools, clinics, or worship before formal subscriber disclosure. Chatrie evaluates the staged acquisition as a search.
What must probable cause establish?
The affidavit should show why the offender likely carried a trackable device, why the provider likely holds the data, and why the defined area and time will probably identify evidence rather than merely sweep innocent users.
Why does the number of innocent users matter?
It affects whether proximity meaningfully indicates involvement and how broad the search is. Population density, residences, businesses, roads, events, and ordinary activity should be disclosed.
What is an accuracy radius?
It estimates uncertainty around a reported coordinate. The actual device may be anywhere within that area. A central dot without the radius can create false precision.
Does a location point prove the person was there?
No. The point estimates a device location. Phones can be shared, left behind, stolen, or carried by someone else. Attribution and corroboration remain necessary.
Can stage two show movements outside the fence?
It can, depending on the warrant and provider process. That broader travel history is a separate intrusion requiring particularized authorization and probable cause under Chatrie’s stage-by-stage framework.
Who decides which anonymous users are identified?
The warrant should set objective rules and preserve judicial control. Undefined officer discretion to “attempt to narrow” creates particularity and neutrality concerns.
Does Google still answer Location History geofence warrants?
Chatrie reports that Google moved Location History storage to individual devices in July 2025 and represented it can no longer respond to that same demand. Other location sources and legacy cases remain relevant.
Are tower dumps the same as geofence warrants?
No. Tower dumps rely on carrier connections to selected cell sites; historical Google geofences used Location History estimates. Both can sweep unknown users, but architecture, precision, records, and legal analysis differ.
What can support a Franks hearing?
Material intentional or reckless misstatements about precision, phone use, population, sensitive locations, provider operation, area, narrowing, or omitted contrary facts can support the required preliminary showing.
Does good faith always prevent suppression?
No. Leon has limits, and Chatrie did not decide good faith. The warrant’s facial breadth, affidavit candor, legal landscape, judicial role, and execution all matter.
What discovery should the defense request?
The complete warrant record, maps, machine-readable coordinates, preservation, provider correspondence, every staged return, native location data, field guides, narrowing criteria, scripts, audit logs, notes, identities, and derivative warrants.
Can evidence from later home or phone warrants be challenged?
Yes. If later process relied on an unlawful geofence stage, derivative-evidence doctrines apply. Build a chronology and excise tainted information from later affidavits.
When should the issue be raised in North Dakota?
Rule 12 generally requires a pretrial suppression motion when the basis is reasonably available. Seek sealed materials and technical discovery early and request leave to supplement if production is delayed.