On June 29, the Supreme Court ruled that when police ask Google to hand over location data for every phone that passed near a crime scene, that request is a search under the Fourth Amendment, the same as searching your house. It's the most significant Fourth Amendment ruling on digital privacy since Carpenter v. United States in 2018, and it got a fraction of the attention paid to the same week's rulings on birthright citizenship and transgender athletes, including in this newsletter, where it was mentioned only in passing. This piece asks the question those other rulings already answered elsewhere: what specifically did the Court decide, and why does it matter more than the coverage suggested?

Chatrie v. United States began with a 2019 bank robbery in Midlothian, Virginia. Police couldn't identify the robber from witnesses or surveillance footage, so they asked a Virginia magistrate for a "geofence warrant" directed at Google: hand over data on every cell phone within a 150-meter radius of the credit union around the time of the crime. Google's records eventually led police to Okello Chatrie.

The Court held, 6-3, that this kind of request is a Fourth Amendment search, meaning police generally need a warrant backed by probable cause to make it. The government had argued that because Chatrie voluntarily let his phone share location data with Google, a private company, he gave up any expectation of privacy in it under a long-standing legal idea called the third-party doctrine (more on that below). The Court rejected that argument, extending its 2018 ruling in Carpenter v. United States, which had already found that historical cell-tower location records held by phone carriers deserve Fourth Amendment protection despite being held by a third party.

Here's the part that matters for what happens next: the Court did not decide whether the warrant police actually used in this case was constitutional. It only decided that a warrant, with the normal Fourth Amendment protections, was required in the first place. Whether this warrant had enough specific probable cause is now back in the hands of the Fourth Circuit Court of Appeals on remand.

Most coverage treated Chatrie as a simple sequel to Carpenter. It isn't, and the difference is the actual news here. Carpenter involved records about one already-identified suspect, pulled from the past, covering 127 days of a single person's movements. It was a rifle shot: police already knew who they were investigating and asked a phone carrier for that one person's history.

A geofence warrant works in the opposite direction. Police don't start with a suspect. They start with a place and a time window, and ask Google to search its entire user base, described in litigation as tens of millions of accounts, for every device that was nearby. Google then returns an anonymized list, police narrow down which anonymized devices look suspicious based on movement patterns, and only then does Google unmask the identities behind the remaining devices. It is a dragnet that starts with everyone and ends with a suspect, rather than a records request about someone police already suspect. Chatrie is the first time the Court has confronted that structure directly, and its holding, that pulling anyone into that pool at all counts as a search, is a bigger doctrinal step than extending Carpenter to a new type of record. It means the initial dragnet step itself needs Fourth Amendment justification, not just the final identification step.

The dissenting justices didn't argue that privacy doesn't matter. Justice Alito, joined by Justice Thomas, argued the Court took the case in the wrong posture and that Chatrie could not show police searched his "papers or effects" under traditional Fourth Amendment principles. Justice Barrett went further on the merits, arguing Chatrie had no reasonable expectation of privacy in data he voluntarily gave to Google in the first place, the classic third-party doctrine position the majority rejected.

Groups like the Electronic Frontier Foundation, which has filed amicus briefs against geofence warrants in multiple cases, argue the technique is a modern general warrant: a tool that searches everyone in an area on the chance one of them is guilty, which the Fourth Amendment was written specifically to prevent. Law enforcement agencies, for their part, have defended geofence warrants as one of the only investigative tools available for crimes with no witnesses or suspects, arguing that without them, some solvable crimes go unsolved. Both sides agree the tool is powerful. They disagree about whether that power should exist without stronger constitutional guardrails than existed before June 29.

The accuracy and error record of geofence data is central to why this ruling matters practically, not just doctrinally. In litigation, Google has disclosed that its location estimates are designed to be accurate within a given radius only about 68 percent of the time, meaning roughly one in three location points placed in an initial search pool may be wrong. That is not a hypothetical risk: Phoenix police arrested Jorge Molina in 2020 for murder based on geofence data placing his phone near the scene, before determining someone else, who had used his car, was the actual suspect. He was cleared, but not before losing his job and spending days in jail. On scale, Google told courts it responded to roughly 9,000 geofence warrant requests in 2018 alone, each one searching a wide swath of its user base to build the initial pool. Those numbers explain why the Court's ruling, that the initial search itself, not just the final identification, requires Fourth Amendment justification, is being read as a bigger deal by privacy lawyers than by the general press.

The third-party doctrine, from Smith v. Maryland (1979) and United States v. Miller (1976), holds that once you voluntarily give information to a third party, like a bank or phone company, you generally lose your Fourth Amendment claim to it. The government isn't searching you; it's asking the third party for records the third party already has.

Carpenter carved out an exception for cell-site location data, for two reasons. Phones generate that data automatically, without any real choice by the user. And the data can reveal a detailed picture of someone's whole life. Chatrie confirms that exception applies to geofence-style location data too, rejecting the government's argument that Google's terms of service change the analysis.

On remand, the Fourth Circuit must apply the ordinary rules for probable cause and particularity, meaning a warrant has to specifically describe what is being searched and why. The question for the court is whether an initial anonymous sweep of an entire area can ever be "particular" enough to satisfy the Fourth Amendment.

The Fourth Circuit's remand decision, with no set deadline, will determine whether the standard three-step geofence process can survive in anything like its current form, or whether police must narrow their initial requests far more than has been typical practice. Other circuits with pending geofence cases will look to that decision. Police departments and prosecutors nationwide are likely already revising how they request this kind of data, seeking narrower time windows and areas, and pairing requests with more independent evidence upfront to support probable cause for the initial pool, not just the final suspect.

There's more room here than in most Fourth Amendment fights. Even before Chatrie, some lower courts, including a Virginia federal court and a California state court, had already found particular geofence warrants unconstitutional as written while leaving open that a sufficiently narrow, well-justified version could pass muster. Chatrie doesn't ban geofence warrants. It requires police to justify the initial search, not just the final identification, the same way they'd have to justify any other warrant. A version of the tool limited to short time windows, small areas, and independent corroborating evidence that the suspect was likely present is very plausibly still constitutional after this ruling. The disagreement isn't over whether police should ever get phone location data. It's over how much specific suspicion has to exist before the search starts.