There is a particular kind of search the Fourth Amendment was written to stop: the kind that begins with a dragnet and ends by deciding who looks suspicious.
In 1791, that dragnet looked like a “general warrant,” a government permission slip to rummage through private papers without naming the person or the place with any real specificity. In 2026, the dragnet looks like a polygon on a map.
The Supreme Court is now weighing whether police can use “geofence warrants” to compel a tech company to identify devices that were within a defined geographic area during a defined time window, even if most of those devices belong to people with nothing to do with a crime. The case is Chatrie v. United States, and it sits at the intersection where constitutional language about “papers” and “effects” collides with a world where our phones keep receipts of our movements.
Join the Discussion
What is a geofence warrant, in plain English?
A traditional warrant is supposed to work like a pointer. It targets a particular suspect, a particular home, a particular device, a particular set of records. A geofence warrant often works like a sieve: it starts by collecting a pool of devices that were nearby, then narrows from there.
In Chatrie, police investigating a bank robbery sought help from Google. They asked Google to search location histories to determine which users were present within the defined “geofence” around the crime scene during the relevant timeframe. Okello Chatrie was later convicted in a prosecution that stemmed from that process.
The constitutional question is not whether police can ever obtain location information. It is whether this method flips the Fourth Amendment’s logic by treating proximity to a crime scene as enough to search the location data of everyone inside an invisible digital perimeter.
The Fourth Amendment problem: is this a modern general warrant?
Chatrie’s lawyer, Adam Unikowsky, framed the government’s tactic in historically loaded terms. He warned the Court that this is the kind of broad, exploratory search the Fourth Amendment was designed to prevent, arguing: There was not probable cause to search the virtual private papers of every single person within the geofence merely because of their proximity to the crime.
That sentence matters because it highlights a core Fourth Amendment requirement: probable cause is supposed to tie suspicion to a person or place, not to a mere coordinate on a map that happens to include hundreds or thousands of innocent people.
If a warrant is meant to be particular, a geofence warrant raises an uncomfortable question: particular about what? A location? A time window? Or the individuals whose data is being searched?
The government’s answer: you gave the data away
The federal government pushed back hard on the idea that this should be treated as a sweeping search. Deputy Solicitor General Eric Feigin warned that Chatrie’s approach would create, in his words, an unprecedented transformation of the Fourth Amendment into an impregnable fortress around records of his public movements that he affirmatively consented to allow Google to create, maintain, and use.
This is the modern version of an old argument: if information is shared with a third party, how private can it be?
But location data is not an ordinary “shared” fact, like handing a bank teller a check. Smartphones generate a continuous log that can reveal patterns: where you sleep, where you worship, who you visit, what clinic you entered, what protest you attended, whether you were at a bar every Tuesday night for three months, and whether your “public movements” are actually private routines no one else was meant to reconstruct.
The justices’ worry: if this is allowed, what else is allowed?
Oral argument made clear that the Court is not only thinking about bank robberies. The hard cases are not hard because anyone doubts law enforcement should investigate violent crime. They are hard because rules built for “the obvious case” tend to be used everywhere else later.
Chief Justice John Roberts pressed the government on the inevitable next step: what's to prevent the government from using this [tool] to find out the identities of everybody at a particular church, a particular political organization?
Feigin answered candidly: I don't think there's any kind of categorical protection around something like a church.
Roberts followed up, sharpening the constitutional stakes: So you don't think there's any constitutional protection from such organizations to be subject to focused surveillance that would cover everybody in a particular location?
Justice Neil Gorsuch distilled what he understood the government’s position to be: you don't think it's a Fourth Amendment search at all. You don't need a warrant.
He then walked through the practical implication, describing the ability to identify everybody who is at a church, or a political rally, or the abortion clinic, or anything else like that
if the government can obtain compliance from a company holding the data.
Feigin’s response was striking in its directness: So I do think that effectively is our answer.
If you want to understand why this case matters, start there. A constitutional right does not mean much if it cannot prevent the government from compiling a list of everyone who attended a disfavored meeting, entered a sensitive medical facility, or gathered to petition for change.
What the Court could do, and what would change
The Court does not have to choose between “police can never use location data” and “police can collect everything.” There are several plausible landing points, and each would reshape Fourth Amendment doctrine in a slightly different way.
1) Declare geofence warrants a Fourth Amendment search that requires a real warrant
This would reject the idea that sweeping up location histories is not a search at all. It would not automatically ban geofence warrants, but it would force courts to treat them as constitutionally serious and demand genuine justification.
2) Allow them, but only with strict limits tied to particularity and minimization
The Court could say: yes, a judge may authorize a geofence request, but only if the request is narrowly tailored, time-limited, and structured to reduce exposure of innocent people’s data. That would import the Amendment’s particularity principle into a world where “place” is digital.
3) Bless them broadly as “reasonable,” with minimal guardrails
This would be the most consequential for everyday privacy. It would tell law enforcement that a judge’s signature on a broad location sweep is enough, even when the initial search begins with everyone and only later narrows to someone.
Justice Ketanji Brown Jackson signaled concern about turning this into an unnecessarily sweeping ruling, telling Chatrie’s lawyer: I see you as making maximalist arguments about this that I'm trying to understand if they're necessary to get to the point that you want to go.
She also suggested focusing on whether the initial geofence warrant was reasonable at the outset, rather than building a larger doctrine around the general-warrant analogy.
That instinct for narrow decisions is familiar at the Court. The problem is that narrow decisions can still build wide pipelines for data.
The deeper issue: the Fourth Amendment’s two languages
The Fourth Amendment speaks in two registers at once.
- One register is concrete: “persons, houses, papers, and effects.”
- The other is principle: the right to be free from “unreasonable searches and seizures,” plus the demand that warrants be supported by probable cause and describe with particularity what is to be searched and seized.
Digital life forces the Court to decide which register controls. If “papers” means only literal paper, privacy shrinks as technology advances. If “papers” means what papers do for human beings, the Amendment can protect modern records that function like diaries, address books, and personal logs, including location histories that reveal the story of a person’s life.
Geofence warrants put pressure on that choice because they do not merely retrieve evidence about a suspect. They search for a suspect by searching everyone else first.
Why this matters even if you have nothing to hide
The familiar line, “If you have nothing to hide, you have nothing to fear,” has always misunderstood the Fourth Amendment. The Amendment is not an award for innocence. It is a structural rule designed to keep government power from becoming casual.
Today, “casual” surveillance does not require a tailing officer or a stakeout. It can be a database query that identifies every device at a location, then converts devices into names, and names into investigative targets.
Even if most geofence searches are aimed at legitimate investigations, the architecture they create is reusable. And reusable surveillance power inevitably drifts toward politics, religion, association, and the kinds of “public movements” that are only public in the sense that your phone was there too.
The question the Court cannot dodge
The Supreme Court can write a narrow opinion. It can decide this case on fine distinctions about consent settings, time windows, or how many steps a company takes before it hands over identities. But it cannot avoid the underlying constitutional question.
Does the Fourth Amendment allow the government to search for the guilty by first searching the innocent?
However the Court answers, it will not just decide a tool for bank robbery investigations. It will announce a theory of privacy for an era where our “papers” are not stored in desks, but in servers, and where our movements are not merely lived, but recorded.