To see why the ruling mattered, you have to start with the rule it strained against, and how much of modern surveillance quietly depends on that rule staying intact.
The third-party doctrine
The Fourth Amendment protects against unreasonable searches, and for most of the twentieth century that turned on whether you had a reasonable expectation of privacy. Two cases in the 1970s built a sweeping exception. In United States v. Miller (1976), the Court held that a person has no Fourth Amendment interest in bank records, because those records are the bank's business documents, voluntarily conveyed. In Smith v. Maryland (1979), it held that the phone numbers you dial are not private, because you knowingly give them to the phone company to connect your calls. A device called a pen register that logged those numbers was therefore not a search.
From these grew the third-party doctrine: if you voluntarily share information with a third party, you lose your reasonable expectation of privacy in it, and the government can obtain it without a warrant. In 1979 that meant a list of dialed numbers. By the 2010s it meant your email metadata, your search queries, your purchase history, and the near-continuous location trail your phone generates by design. The doctrine written for a pen register had quietly swallowed the digital record of a life.
The facts of Carpenter
Timothy Carpenter was convicted as the organizer of a string of armed robberies of electronics stores around Detroit in 2010 and 2011. To place him near the crime scenes, the FBI obtained his cell-site location information, the records of which cell towers his phone connected to, from his wireless carriers. Prosecutors got 127 days of that data, mapping thousands of location points across roughly four months.
They did not get a warrant. Instead they used a court order under the Stored Communications Act, which required only "specific and articulable facts" that the records were relevant to an investigation. That is a materially lower bar than the probable cause a warrant demands. Carpenter argued the acquisition was a Fourth Amendment search that required a warrant, lost in the lower courts on third-party-doctrine grounds, and the Supreme Court agreed to hear it.
What the Court actually held
Chief Justice Roberts, writing for the majority, drew a line around cell-site location data. The third-party doctrine, he reasoned, was not built for a record this comprehensive. Location data is generated automatically, without any affirmative act by the user, and it accumulates into something qualitatively new.
"In light of the deeply revealing nature of CSLI, its depth, breadth, and comprehensive reach, and the inescapable and automatic nature of its collection, the fact that such information is gathered by a third party does not make it any less deserving of Fourth Amendment protection." Chief Justice Roberts, majority opinion, Carpenter v. United States (2018)
The Court held that acquiring historical cell-site records is a search, and the government generally needs a warrant supported by probable cause to get them. The majority leaned on the same intuition the concurrences in an earlier GPS-tracking case had raised: a detailed, retrospective map of where a person has been reveals their familial, political, professional, religious, and sexual associations, an intimate window that no single call record ever opened.
Roberts stressed the decision was "a narrow one." It did not overturn Smith or Miller. It did not touch conventional surveillance tools, business records generally, real-time cell-site tracking, or foreign-intelligence gathering. The Court declared cell-site location a special case and pointedly declined to say how far the reasoning extends. That narrowness is why the fight over the doctrine is far from settled.
The limits, and what stayed open
Because Carpenter was written so tightly, a great deal of surveillance sits just outside it. Courts and agencies are still contesting where the edges fall:
- Shorter windows. The Court reserved judgment on whether a brief period of location data, rather than 127 days, triggers the same protection.
- Tower dumps. A request for every phone that connected to a given tower during a window sweeps in thousands of uninvolved people, and its status remains unsettled.
- Real-time tracking. Carpenter concerned historical records; live location tracking was explicitly left for another day.
- The data-broker workaround. The ruling constrains what the government can compel from a carrier. It says much less about the government simply buying location data on the open market. Agencies have purchased app-derived location feeds from commercial data brokers, an approach that sidesteps the warrant requirement Carpenter created.
That last point is the sharpest. A warrant rule that applies when police demand your records from AT&T, but not when they buy the same movements from a broker who harvested them through a weather app, has an obvious hole. Several legislative efforts have tried to close it by barring agencies from purchasing data they would otherwise need a warrant to obtain. The gap remains a live policy question.
Why encryption outlasts the legal fight
Carpenter is a genuine limit on government power, and it matters. It is also fragile in a way worth being clear-eyed about. It protects a specific category of data, it can be narrowed or worked around, and it governs only what US courts can reach. Legal protections shift with the composition of a court and the text of a statute. They are worth defending, and they are not a foundation you would want to stand on alone.
There is a more durable form of protection, and it is architectural rather than legal. Data that is never collected cannot be subpoenaed, purchased, or leaked. Content that is end-to-end encrypted cannot be handed over in readable form, whatever the legal standard, because the provider never holds the keys. This is the difference between metadata surveillance and content surveillance, and it is why data minimization is the strongest privacy tool a service can offer. A company that structurally cannot see your messages has nothing to turn over when the order arrives.
Haven is built on that principle, alongside the many other tools and services working the same problem. The point is not that law does not matter. It is that legal and technical protections cover different failure modes. Carpenter guards the front door where the state knocks on a carrier's records. Encryption and minimization make sure that even when a door opens, there is less behind it worth taking. You want both, and you should not mistake one for the other. For the wider picture of how these forces interact, see our overview of the gaps in electronic surveillance law.