The Supreme Court is currently weighing a case that could reshape how law enforcement works with technology firms to identify potential suspects. In the next few weeks, the court is expected to rule on whether or not so-called geofence warrants are legal under the Fourth Amendment, which prohibits unreasonable searches and seizures and sets requirements for issuing warrants.

Lawyer Adam Unikowsky argued the case last month on behalf of petitioner Okello Chatrie, who was charged with robbery after law enforcement obtained a geofence warrant directed at Google. Under such warrants, tech companies are asked to give law enforcement the location history of people in a certain area so they can identify unknown suspects, and Google has said it has been served with geofence warrants covering exceptionally large areas across multiple days.

The case is the first involving digital privacy to hit the Supreme Court since 2018 and could have major implications for other types of police tools that involve large amounts of data. Unikowsky spoke with Recorded Future News about why he believes geofence searches are problematic, how people do not understand what they are consenting to and why the way the court rules could have a dramatic impact on Americans’ right to privacy.

This interview has been edited for length and clarity. 

Recorded Future News: The Chatrie case is the first digital privacy case, as well as the first Fourth Amendment case, to hit the Supreme Court since a decision in 2018 that limited police use of historical cell-site location data. Why do you see the Chatrie case as important?

Adam Unikowsky: The Supreme Court hasn't heard a case about digital privacy in several years, and the Carpenter case — which you referred to from 2018 — left a lot of questions unanswered. It left questions about what constitutes a search that requires a warrant unanswered and it also left questions unanswered about exactly what types of warrants could be issued. 

The Chatrie case involves a novel type of warrant called the geofence warrant, which is a type of so-called reverse warrant. It started out in a situation where the police did not have any suspects and used the warrant in order to find the suspect as opposed to a traditional warrant which focuses on searching a particular suspect. The case presents a lot of issues that prior Supreme Court cases haven't addressed, but which are important to modern law enforcement techniques. 

RFN: Some legal scholars believe that depending on how the court rules, it could end up leading to a surge in reverse searches — for example, reverse keyword searches where police can see everyone who has searched web browsers for specific keywords. 

AU: The court's ruling could address the constitutionality of reverse keyword searches as well as reverse AI searches. In other words, searches in which the police check every single person's Google account as to which people search for a particular search term that seems relevant to a crime that they're investigating or search everyone's AI chats for questions that are probative of whether a person has committed a particular crime. One of the issues in the case is whether this type of warrant — which allows everyone's accounts to be searched — is consistent with the Fourth Amendment, and if so, exactly how broadly or how narrowly do such warrants have to be drawn.

RFN: The Fourth Amendment and digital privacy legal scholar Orin Kerr said he believes the justices will likely conclude that geofence warrants can be drafted constitutionally, but also will say that they have to be limited in time and space. Do you agree with that? 

AU: I'm not going to speculate on how the court will rule on the oral argument. I will say that we put a number of arguments on the table. One of those arguments was that the geofence warrant was a general warrant because it impermissibly authorized the police to search every person's account. We've also put narrower arguments on the table, such as the fact that the geofence was overly broad because it encompassed a bank and the church for a one hour stretch and there wasn't probable cause to search what we understand to be the virtual private papers of every single person in the church and every single person in the bank merely because of their proximity to the crime. 

RFN: If the justices set up strict parameters for how specific geofence warrants need to be in the future, is that a victory? The geofence warrant in this case covered a 150-meter radius and roughly 17.5 acres. In its amicus brief, Google pointed out that it has gotten geofence warrants for areas as big as one that covered 2.5 square miles of San Francisco for more than 48 hours of time. 

AU: Ultimately, our goal is for the court to hold that there was a Fourth Amendment violation. We have several arguments. I do think that those sorts of very broad warrants are very troubling. So if it were the case that the court were to hold that geofence warrants are, in some cases, constitutional — and I'm not going to speculate as to whether they will or will not reach that conclusion — I would favor a rule under which any geofence warrants would have to be drawn narrowly and not so broadly as to encompass potentially thousands of people.

RFN: Justice Alito pointed out that Google no longer stores location history and can no longer comply with these warrants. But several other websites and apps do collect and store location history. Is it moot that Google no longer stores it because this ruling would apply to all sorts of other online services?

AU: The opinion would have wider implications than just Google. This particular case involved the search of Google servers, but the collection of location information is a ubiquitous practice. This case attracted a lot of attention from the tech community. There were nearly 30 amicus briefs filed, so obviously a lot of people think that this is an important and live issue, regardless of Google's own business practices. 

RFN: You argued that the geofence warrant in this case qualifies as a general warrant, and those are unconstitutional. But Justice Sotomayor said she disagrees because it identified a place, a crime and a limited timeframe. Why do you believe it is a general warrant and why are general warrants unconstitutional?

AU: I'm not going to speculate on Justice Sotomayor’s view on the case, but I'll just explain the argument that we were making. A general warrant is a warrant that allows the police to search everybody for evidence of a crime, even people who aren't suspected of committing a crime. 

This is a practice that existed before the Declaration of Independence that was truly loathed by the colonists. The police would develop suspicion that people weren't paying taxes and they would just come into every single person's house and try to find goods on which taxes weren't paid or they would rummage through every person's papers without individualized suspicion. A general warrant is a term for a warrant that is not particularized, that doesn't name the place to be searched and the things to be seized. It just allows general authority to the police to rummage through people's possessions. 

In this case, one of the arguments we advanced is that the police are effectively searching every single person's Google account to figure out who was within the particular area. The way these searches work is that the police go to Google and tell Google, ‘Hey, we want you to figure out all the people who were within the geofence,’ which is a search near the time and place of the crime. Everyone has a Google account that stores their location data in their own account so there's no way for Google to identify the people who are near the scene of the crime unless they search every single person's account to find location history that's within the geofence. 

We understand the warrant to search every single person’s Google account, and we're talking about tens of millions here, is akin to a general warrant, which authorized the search of thousands of houses for evidence. This argument hinges on two points. One is that Google, when it does the search, is acting as an agent of the government, which we think it is because it's carrying out a search warrant. Second, peering inside someone's account is equivalent to a search of the account. 

RFN: It seemed like some of the justices agree with your argument that if location history searches are allowed, then searches of all kinds of things stored in the cloud, like photos, emails and documents, could be allowed. Why do you think that if the court allows geofence warrants, it could extend to other content, like photos, documents and emails?

AU: Several of the arguments the government was making would have applied with identical force to emails and documents. The government relied on language in Google's privacy policy, for example, saying that Google reserves the right to respond to search warrants, and that Google, in some cases, could disclose [user data] for public safety purposes. The government interpreted some of those statements in the privacy policy to imply that people had essentially waived any interest in the data, and therefore the government's accessing the data on Google servers just wasn't a search at all. 

But those privacy policy statements apply to all data stored in Google, not just location history, It applies to email as well and documents and photos and calendar entries so, as a result, many of the government's arguments really would imply that there's no protection to data in the cloud at all.

RFN: The government has said the millions of people searched remain anonymous, but you argued that's somewhat illusory because one's movements within the geofence often function as a kind of fingerprint. 

AU: One of the arguments the government made was that people aren't really being searched unless the police know subjectively who they are. So if the police get a list of people's movements in the geofence — like the person who was in the bank, the person who went to the church or the person who entered the bank and left the bank — but they don't know who those people are yet, they haven't really been searched because there's no privacy interest in [police] just seeing anonymous movements. 

We argued that it doesn't matter if they're anonymous — the police still can't look at their possessions. For example, if the police go into your house and read your diary that's a search of your house. Even if the police don't know who owns the house, they still read your diary. We understand that viewing people's movements is the same thing. 

The second argument we advanced is that anonymity is quite illusory. Our expert witness in this case inferred the actual identity of three of the people within the geofence from their movements over a two hour stretch based solely on their movements and public records. It’s often very possible for police to take just one or two hours worth of someone's movements, look at them carefully, compare those to public records and figure out who they are. The idea that anonymity really protects people's privacy is not realistic.

RFN: The government argues that people give this data to Google voluntarily and that if they read the privacy policy, they'll know that they're consenting. 

AU: When you're setting up your phone, you'll just get some vague message saying ‘we might share your location.’ You click yes and you may not realize what you're doing. In this case, Mr. Chatrie used an Android phone. Turns out, when you're setting up Google Assistant on your Android phone, there's language that says that the Google Assistant might not work correctly unless you enable a bunch of services, one of which is location history. People set up their phones and get these messages that their phone won't work unless they click yes, so they click yes. 

People may not have a full understanding of what they're doing and they certainly don't understand that they've suddenly consented to all of their movements being shared with the government. It's true that very general descriptions of location history are provided like we'll keep a record of where you go, but people are not told that every two minutes Google will know their location within three meters and keep it potentially forever and give it to the government if the government asks for it. That level of granularity is not disclosed. It's just not realistic that people understand that when they click yes at 2 a.m. when they're trying to get their phone to work.