Jake Laperruque is a Tech Policy Press Fellow and the Deputy Director of Security & Surveillance at the Center for Democracy & Technology. This piece is written in his personal capacity for the fellowship and does not necessarily reflect the views or positions of his employer

This week, the Supreme Court heard oral arguments in Chatrie v. United States, a Fourth Amendment case with potentially profound implications for location tracking and privacy in the digital age. The case involved the use of a geofence warrant — a novel surveillance technique that first emerged and has since skyrocketed in scale over the past decade — in which police demand information on all cellphones within a “geofence,” a set area and period of time, often based on the place and time where a crime occurred. Although predicting the outcome of a case from oral argument can sometimes be a fool’s errand, the discourse during argument offers interesting insights into how this case may play out on numerous important issues.

Is it even a search when police demand location data that’s “voluntarily” shared with a digital service?

Fourth Amendment law generally holds that if private information is voluntarily shared with a third party, it is no longer entitled to protections from government search and seizure (a concept called the “third-party doctrine”). So it’s no surprise Justices questioned whether using a digital service that collects location data — in this case Google’s Location History — meant that geofence warrants did not actually involve a Fourth Amendment search.

However, the Court seemed reluctant to embrace this view. A plethora of examples were raised in which relying on a third party doesn’t appear to negate Fourth Amendment protections. These ranged from analog systems like safe deposit boxes and the postal services, to digital services like cloud storage for photos and documents. At one point, Justice Kagan asked if services would have to meet a “nobody-can-live-without-this standard” to avoid triggering the third-party doctrine, seeming skeptical of setting the bar so high. While Justice Kavanaugh seemed among the most reluctant of the Justices to strike down the geofence warrant, he framed questions around the notion that a search had occurred, but that the warrant at issue was proper. Even the government conceded that data users share via email and other third-party digital services likely couldn’t be treated as voluntarily shared and freely grabbed by police.

If the Supreme Court does indeed rule that demanding location data from a service provider is a search, it would be an important victory for privacy rights in a digital age. But it would still leave major questions about how broad the parameters of such searches can be.

Are geofence warrants inherently overbroad?

One argument raised by the appellant is that geofence warrants constitute a “general warrant” — a dragnet that by its nature is unconstitutionally broad — because determining which phones were within the geofence required running a scan through Google’s full database to pull up relevant records. While this theory prevailed in a separate geofence case before the Fifth Circuit Court of Appeals, it appeared unlikely to carry weight with the Supreme Court. Justices pushed back against this concept, noting that software, rather than humans, conducted the scan and that automated searches of databases occur across many systems. Overall, the Court seemed less interested in categorically rejecting geofence warrants and more focused on what the proper bounds of a geofence could be.

How broad can geofences be?

Justices focused significantly on how broad geofences could be and how bounds could effectively be set on them - these could be the decisive question for how the case will play out.

Allowing excessively expansive geofences raises serious risks. Google’s amicus brief highlighted shockingly broad geofence warrants that spanned hundreds of acres or even entire sections of cities. It's not uncommon for geofences to sweep in sensitive areas such as a house of worship (the geofence warrant in Chatrie included a church). And as several Justices noted, geofence warrants aren’t limited to public spaces – they can catalog movements inside homes and other private areas.

It may be possible to place tighter limits on geofence warrants both in terms of place and time. Plaintiff’s counsel raised the concept of allowing a geofence if the suspect is the only person whose data would be collected, with no extraneous surveillance of others (such as an isolated area that only a shooter and victim were in). But this received pushback from the Court, with Justice Kavanaugh highlighting Court precedent that searches could be permitted based on probable cause to believe evidence would be found, even if the search wasn’t of a suspected criminal. On the same point, Justice Gorsuch asked if the geofence in Chatrie could have been limited to the bank where the robbery being investigated occurred, as opposed to the broader 17.5-acre geofence that was used, inquiring, “There might be probable cause to think that everybody in the bank might have some information?”

Could the Court say geofences are permissible if there is probable cause that everyone within it would be either a perpetrator or a witness to a crime? Such a ruling that geofence warrants need to be tailored to the specific area where a crime occurred — pulling in only those with unique knowledge of the crime — could appeal to Justices as a way of allowing geofence warrants but drawing a line that stops them from becoming sprawling fishing expeditions.

Will the Court just kick the can on core constitutional questions?

Another issue that received attention in arguments was the good-faith exception to the exclusionary rule. Even if a search is unconstitutional, the fruits of that search can generally still be used as evidence if police reasonably believed that the search was lawful. Justices Alito and Thomas suggested that the Court could and should simply rule on those grounds — upholding the search without addressing any of the substantive questions — with the former opining that the case was inviting a “law review article” rather than a necessary legal ruling.

This would be a confusing response to an issue that has split the lower courts (leaving different rules active across the country), as well as a huge waste of an opportunity to address important, forward-facing privacy issues. Perhaps most importantly, it could also create a calamitous Catch-22 for privacy rights: If courts refuse to set rules on novel surveillance techniques because police acted under a good faith belief they were lawful, police would be able to exploit legal limbo where they can reasonably believe future uses of those techniques are lawful (which in turn would again be shrugged off by courts without addressing the technique’s legality).

A related issue some Justices raised was whether changes in data policies have rendered this issue moot. In late 2023, Google began a (now completed) shift in its data management policies: location data is now stored on users' devices rather than on the company’s servers, and as a result, Google is no longer able to provide responsive data for geofence warrants. But while Google is the primary known recipient of geofence warrants, other companies will inevitably receive them as well (and likely already have). It would be shortsighted to kick the can on this case because Google’s practices have changed, only to repeat the process as soon as a geofence warrant to another company is challenged.

What other technologies and tools could this ruling apply to?

The Court’s ruling in Chatrie could have ramifications far beyond geofence warrants. On the one hand, it could broadly establish safeguards for data held in the cloud. On the other hand, it could open the door to a wide variety of “reverse warrants” in which sensitive data is seized for a group of people rather than an individual suspect.

But the Court seemed hesitant to engage on these topics. Throughout arguments, the Justices seemed to take for granted the notion that data held by common cloud-based services (like email and online calendars) were entitled to warrant protections despite being voluntarily shared with a third party. However, they spent little time examining what overall doctrine might shape this view, and potentially replace the third-party doctrine for digital services. (Perhaps the closest guiding point came in a colloquy between Justice Kagan and plaintiff’s counsel, when he suggested a clear and affirmative consent standard, where “it's very clear from the interaction that you really are consenting to the other person doing whatever they want with it.”)

And while one other type of reverse search — “tower dumps,” a location surveillance technique similar to geofence warrants but far less precise —was mentioned, it was discussed only briefly, with the Court seeming intent on separating them from this case. Other reverse searches — such as reverse keyword warrants — were not raised at all. Perhaps most eyecatching was the Court’s fleeting reference to the data broker loophole, with Justice Barrett asking about the implications of police simply buying data rather than obtaining it via a warrant. This issue has become one of the top surveillance reform issues for privacy advocates in Congress, and if it is not addressed by legislation, it is all but certain to come up before the Court eventually.

But given the limited focus they received, resolution of these issues may have to wait for another day as the law continues to struggle to keep pace with technology.