You carry your phone. Your phone talks to cell towers. Your carrier keeps the logs.
For years, that basic infrastructure quietly reshaped Fourth Amendment law. Not because the Constitution changed, but because the most revealing “search” in modern life often looks like paperwork. A request. A database. A file handed over by a company that is not you.
In Carpenter v. United States (2018), the Supreme Court confronted a question that sounds technical but is actually intimate: when the government collects your historical cell-site location information, is it doing a “search” under the Fourth Amendment? And if so, does it need a full warrant supported by probable cause?
The Court’s answer was a pivot for the digital era: yes, this is generally a search, and yes, police generally need a warrant, even though the data sits with a third party.
One more detail matters for how people cite Carpenter: the decision was 5 to 4, written by Chief Justice Roberts, and it treated 7 days or more of historical CSLI as the kind of “extended period” that triggers Fourth Amendment protection. (Carpenter’s case involved 127 days of records.)
Join the Discussion
The Fourth Amendment issue
The Fourth Amendment protects “the right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures.” Traditionally, that meant physical intrusions: opening a desk drawer, entering a house, patting down a pocket.
But location data does not live in drawers. It lives in logs.
CSLI is created when your phone connects to nearby cell towers as it communicates. That includes calls and texts, but also data sessions and background network signaling. The result is a time-stamped record that can be used to infer where you were, and where you went, over days and weeks.
If you want to picture it, think: tower A at 10:32 p.m., tower B at 10:41 p.m. One line is not much. A long series becomes a trail.
In Carpenter, the government obtained a long stretch of these historical records from wireless carriers without a traditional warrant.
What the government did
Police investigating a string of robberies sought Timothy Carpenter’s historical CSLI from his cell phone providers. Instead of applying for a standard Fourth Amendment warrant, they relied on the Stored Communications Act, which allowed a lesser court order under 18 U.S.C. § 2703(d).
That difference matters. A warrant requires probable cause and is issued by a neutral judge. A § 2703(d) order can issue based on “specific and articulable facts” showing “reasonable grounds to believe” the records are relevant and material to an ongoing criminal investigation.
The CSLI placed Carpenter’s phone near multiple robbery locations, and the records became key evidence at trial.
Why third-party possession did not end it
The most important friction point in Carpenter was the third-party doctrine, a line of cases that, at a high level, says this: if you voluntarily share information with a third party, you generally assume the risk that the government can obtain it from that third party without a warrant.
Older examples include bank records and numbers dialed on a telephone. If the company has it, and you conveyed it, the argument goes, your reasonable expectation of privacy is reduced.
Carpenter did not fully overthrow that doctrine. But it refused to let the doctrine swallow location privacy whole.
Why CSLI is different
- It is deeply revealing. Long-term location data can sketch a person’s life, including visits to religious services, doctors, political meetings, romantic partners, and private homes.
- It is effectively unavoidable. Carrying a phone is not a niche choice in modern life. Location records are generated by the network as a byproduct of participation in society.
- It is comprehensive over time. A single data point is one dot. Days and weeks of dots become a map of habits, associations, and patterns.
So even though the records were held by wireless carriers, the Court treated the government’s acquisition of extensive historical CSLI as a Fourth Amendment event. The constitutional question was not only who stored the data, but what the data revealed and how it functioned as surveillance.
The holding
The Court held that accessing historical CSLI for an extended period is a search under the Fourth Amendment, and that the government generally must obtain a warrant supported by probable cause to get it.
For practical purposes, Carpenter’s headline metric is that the Court treated 7 days or more of historical CSLI as enough to trigger the warrant requirement. The case itself involved 127 days of location records.
This is the core shift Carpenter represents: digital records can be constitutionally sensitive even when they sit in corporate hands.
How narrow was it
Carpenter was written as a digital-era opinion with guardrails. The Court emphasized that its ruling was narrow, aimed at historical CSLI and the privacy risks of long-term tracking.
In practical terms, Carpenter leaves several questions open or only partly answered, including:
- Real-time tracking. Carpenter addressed historical CSLI, not live or prospective CSLI.
- Short time windows. The Court drew its line at 7 days for purposes of the case, but did not definitively resolve whether a shorter request is a “search” in the same way.
- Other business records. The Court did not announce that all data held by third parties now requires a warrant.
- Traditional emergencies. The Court signaled that exigent circumstances can still justify warrantless action when lives or safety are at stake.
It also did not squarely decide how the Fourth Amendment applies to newer tactics like tower dumps, geofence requests, or government access to brokered location databases, even though those issues hover in the background of modern investigations.
Contrasts with older doctrines
Third-party doctrine
Before Carpenter, the government often argued that records held by companies were easier to access constitutionally because the individual had exposed them to a third party.
Carpenter did not erase that idea, but it carved out a significant exception: certain categories of digital data are so revealing and so involuntarily generated that treating them like ordinary business records is constitutionally inadequate.
Search incident to arrest
Another older doctrine lets police search an arrested person and the immediate area without a warrant, largely for officer safety and to prevent destruction of evidence.
In the phone era, the Court has been skeptical of extending that logic to the contents of digital devices. In Riley v. California, for example, the Court held police generally need a warrant to search a cell phone seized incident to arrest. (A physical pocket search is not the same as downloading the story of your life.) Carpenter fits that broader theme: digital scale changes the privacy stakes, so analog shortcuts do not automatically carry over.
Why Carpenter matters
Carpenter is not just a “phone case.” It is a clue to how the Court is willing to think about privacy when surveillance becomes frictionless.
The opinion suggests a constitutional instinct that becomes more urgent each year: when technology enables the government to reconstruct a life at low cost and high confidence, the Fourth Amendment has something to say, even if the evidence arrives through a corporate intermediary.
At the same time, Carpenter’s narrowness is part of its story. The Court moved, but it moved carefully, leaving lower courts to fight over the boundary lines in future cases involving app data, geofences, brokered location databases, and other forms of digital exhaust.
Glossary
- CSLI (Cell-Site Location Information): Records created when a phone connects to a carrier’s cell towers, showing which towers handled a device’s communications at specific times and allowing location to be inferred over time.
- Stored Communications Act (SCA): A federal statute that includes procedures for government access to certain stored customer communications and records, including § 2703(d) orders used in Carpenter.
- Warrant: A court order authorizing a search or seizure, typically requiring a showing of probable cause and specifying what can be searched or obtained.
- Probable cause: A constitutionally required level of suspicion for most warrants, based on facts suggesting evidence of a crime will be found in the place to be searched or the data to be seized.
- Carpenter rule of thumb: If the government wants 7 days or more of historical location tracking through CSLI, it generally needs a warrant, even though the records are held by a phone company. The Court did not fully resolve shorter requests.
The question it leaves
The Fourth Amendment was written for a world of doors, desks, and locked trunks. Carpenter is what happens when a Court admits that the modern equivalent is a server.
The deeper question is not whether your phone company has your location records. It does. The question is whether the Constitution allows the government to scoop up that history with something less than probable cause and call it ordinary.
Carpenter’s answer was cautious but clear: not anymore, at least not for extended historical CSLI.