Hook
The tower pinged at 11:47 PM. Your phone was there. The robbery happened at 11:52. Now you’re on a list.
Not because anyone was watching you. Because a detective drew a circle on a map, named a time window, and asked the phone company: who was inside it? Your phone was. So was everyone else’s within three blocks. Forty-seven devices. Forty-seven names. One of them became a suspect. The others went into a file somewhere.
The Supreme Court is deciding whether that’s a search under the Fourth Amendment — whether police need probable cause before they pull that list, or whether proximity is enough. How does being near something become evidence you did something?
How Location Becomes A Suspect Pool
A geofence warrant works like this: police give a service provider — Google, Apple, a cell carrier — a geographic boundary and a time range. The provider runs a query against its location database and returns every device ID that entered that boundary during that window. Police narrow the list, sometimes with a second query. Then they ask for names.
The first query is anonymous. Device IDs only. The second reveals who. This is supposed to be the privacy protection. It isn’t. The device ID is tied to an account. The account is tied to a person. The anonymity is procedural, not real.
This isn’t targeted surveillance. You’re not on a watchlist. There’s no warrant with your name. The warrant names a place and a time. You become visible because your phone was there.
Cell towers log every phone that connects. Not GPS-precise — tower range is blocks, sometimes miles. Your phone connects to the nearest tower, or the least congested one, or the one with the strongest signal. The tower records: this device, this time, this direction. The data sits in a database. The warrant opens the database.
In 2020, Google reported it received more than 11,000 geofence warrants. That number has grown. The tool works. Investigators say it’s often the only lead they have. A robbery in a commercial district at 2 AM — no witnesses, no cameras pointing the right direction, no suspect. The phones were there. The phones know.
The Fourth Amendment says searches require probable cause. The question before the Court: is opening that database a search? Or is it just asking who was nearby?
When Near Becomes Suspected
Your phone pings the tower you’re near, not just the tower where you are.
A man in Virginia was charged with a series of robberies. His phone pinged near six robbery locations over three months. He wasn’t at the robberies. He lived in the neighbourhood. He drove past those intersections regularly. His morning commute took him through the geofence. His phone logged it. The algorithm surfaced his name.
Prosecutors said: six times is not coincidence. The defence said: he lives here. The jury saw the map — six pings, six crimes, same device. He was convicted. The conviction was later overturned on other grounds, but the location evidence was admitted. The judge ruled the geofence warrant was lawful.
Tower range matters. In dense urban areas, one tower might cover two blocks. In rural areas, five miles. When police draw a geofence around a crime scene, they’re drawing it around the tower’s coverage area, not the building where the crime happened. Everyone whose phone was in that coverage area during that time window is in the return set.
You don’t know you’re in it. There’s no notification. The warrant is sealed. The query runs. Your device ID comes back. If you’re eliminated in the first pass — wrong direction, wrong movement pattern — you’re never named. If you’re not eliminated, your name is requested, and then you’re in the file. You learn about it when a detective shows up, or when you’re charged, or never.
Being in the geofence means you were somewhere in that radius during that time. Not at the place. Near it. The margin of error is the margin between suspect and bystander. The system doesn’t distinguish. It returns everyone, and then investigators decide.
What Changes If The Court Says No
If the Court rules geofence warrants require probable cause, police would need a specific suspect before pulling location data. They couldn’t use location data to generate the suspect.
The data still exists. Every phone still pings. Every tower still logs. Every company still stores it. Some for weeks, some for months, Google indefinitely if you have location history turned on. The database doesn’t shrink. The question is: who decides when opening it is a search?
Prosecutors say geofence warrants solve cases that couldn’t be solved otherwise. A shooting in a park at night, no cameras, no witnesses willing to talk. The phones were there. Pull the list, cross-reference against gang databases or prior arrests, interview the names that flag. Without that tool, the case stays cold.
Civil liberties groups say that’s exactly the problem. The warrant doesn’t name a suspect. It names everyone, then makes them suspects. The Fourth Amendment was written to prevent general warrants — the king’s men searching every house on a street to find the one they wanted. A geofence warrant is a general warrant with a timestamp.
The Court heard arguments this week. The question they’re deciding is narrow: does pulling location data for everyone in a geofenced area constitute a search that requires probable cause, or is it just a request for records? Narrow legally. Wide practically. The answer determines whether the starting point of an investigation can be a database of where everyone was.
Cell carriers say if the ruling goes against them, they’ll still have to comply with warrants — just more specific ones. Google says the same. They’re not fighting the warrants. They’re trying to clarify what makes one lawful. The number of requests hasn’t decreased. The boundaries drawn haven’t gotten smaller. What’s changed is that the tool is now visible enough to be challenged.
Close
Your phone is still pinging. The towers are still logging. The database is still growing. The Court will decide whether being findable is the same as being tracked.