The Surveillance Grid You Never Voted For: Flock Cameras, ICE, and the Courts’ Failure to Catch Up

Sometime in the past few years, a camera appeared on a pole near your home. You probably did not vote on it. Your city council may have approved a contract you never heard about. The company that makes the camera — a Georgia-based firm called Flock Safety — now operates an estimated 80,000 license plate readers across more than 5,000 communities in 49 states. And the data those cameras collect, recording every vehicle that passes with a timestamp and location, is being shared with Immigration and Customs Enforcement.

No warrant required. No community consent. In most places, no meaningful judicial oversight at all.

What Flock Does — and Who It Tells

Flock Safety markets its Automated License Plate Reader (ALPR) cameras primarily to local police departments and homeowners associations, promising to reduce crime by logging every passing vehicle’s plate, make, model, and color. The pitch works: the company has grown explosively and now counts hundreds of law enforcement agencies among its clients.

What the marketing materials do not lead with is the data-sharing architecture underneath. Local Flock networks are stitched into a nationwide intelligence fabric. When a local police department signs a Flock contract, it typically also joins a network that allows other agencies — including federal ones — to query plate sightings across the entire grid. ICE has used this network to locate and arrest immigrants, including people with no criminal record beyond their immigration status, in jurisdictions whose elected officials believed they had adopted sanctuary policies.

The mechanism sidesteps sanctuary laws neatly: the local police department is not calling ICE. It has simply joined a data-sharing arrangement that makes its camera feeds visible to federal databases. The political promise of sanctuary dissolves into a technical footnote.

The Fourth Amendment Problem

The constitutional question is not subtle. The Fourth Amendment prohibits unreasonable searches and seizures and, under a long line of cases, generally requires law enforcement to obtain a warrant based on probable cause before conducting targeted surveillance of a person.

Flock’s network does not work that way. It conducts continuous, suspicionless surveillance of every vehicle in a jurisdiction and retains that data — sometimes for thirty days, sometimes longer depending on local contract terms — available for retrospective querying by any authorized agency. There is no prior judicial approval. There is no individualized suspicion. There is just a permanent record of where your car has been, available to be searched whenever an agency decides to look.

The Supreme Court’s 2018 decision in Carpenter v. United States is the most relevant precedent, and it cuts against the government’s comfort with this arrangement. Chief Justice Roberts, writing for a 5-4 majority, held that police must obtain a warrant to access historical cell phone location data from a carrier. The Court reasoned that comprehensive location tracking over time — even of data held by a third party — implicates a reasonable expectation of privacy because it “achieves near perfect surveillance” and can reveal the “privacies of life.” A week’s worth of cell location data was enough to trigger the warrant requirement.

ALPR data from a network like Flock’s is structurally identical. It is comprehensive, time-stamped location data retained over time and available for retrospective government search. The only difference is that it tracks vehicles rather than phones — a distinction that should be legally irrelevant to Carpenter’s core logic.

And yet courts have been slow to apply Carpenter to license plate readers. Civil liberties litigants have raised the argument, but the case law remains thin and inconsistent. The third-party doctrine — the pre-Carpenter rule that information voluntarily shared with a third party receives no Fourth Amendment protection — still exerts gravitational pull on lower courts confronting surveillance technology. Some courts have found that driving on a public road and being recorded by a camera on a public pole involves no reasonable expectation of privacy. Carpenter did not cleanly abolish this reasoning; it carved an exception to it and left the boundaries uncertain.

That uncertainty is where mass surveillance currently lives.

The Grassroots Circuit Breaker

While courts deliberate, some Americans have arrived at a more direct solution. As Brian Merchant reported in Blood in the Machine, people across the country have been physically dismantling Flock cameras — cutting them down, spray-painting lenses, and in some cases removing them from their poles entirely. These acts of sabotage have occurred in multiple states and appear to reflect a loose, decentralized movement of people who have concluded that the legal and political systems have failed to check the surveillance state fast enough.

The phenomenon is worth taking seriously not as a model for civic action — destroying government or contractor property carries serious criminal consequences — but as a signal. When people begin dismantling infrastructure with their hands, it usually means they have decided that no legitimate channel will hear them in time. The cameras go up overnight, the contracts are signed without public hearings, the data flows to ICE before anyone can file an injunction, and by the time a lawsuit works through the courts, the surveillance grid has expanded by another ten thousand cameras.

The gap between the speed of surveillance deployment and the speed of democratic or legal response is not an accident. It is a feature of how this technology gets sold and adopted.

What Courts Could Do — And Have Not

The Supreme Court, to its credit, gestured at the right answer in Carpenter. Roberts wrote that the Court must be sensitive to “seismic shifts in digital technology” that allow the government to conduct pervasive surveillance that would have been unimaginable to the founders. The majority explicitly declined to extend the third-party doctrine to comprehensive location tracking.

But the Court left significant work undone. It expressly limited its holding to the facts before it — 127 days of historical cell site records — and declined to say how far the warrant requirement extends. License plate readers, which operate in real time and in public space, present a harder doctrinal case than cell data, and the Court has not yet taken a case squarely presenting the question.

Lower courts and state legislatures have partially filled the gap. A handful of states have passed laws restricting ALPR data retention periods or limiting data-sharing with federal immigration authorities. California, for instance, has enacted restrictions on how long ALPR data may be retained and who may access it. But these protections are patchwork, unevenly enforced, and easily circumvented by the multistate network architecture that companies like Flock have deliberately built.

Congress has not acted. Federal legislation establishing a baseline warrant requirement for ALPR data does not exist.

The Accountability Vacuum

What makes Flock’s ICE data-sharing so difficult to challenge legally is the deliberate fragmentation of accountability. The local police department did not call ICE. Flock did not call ICE. ICE queried a network database that a local department contributed to under a contract the department signed. Responsibility is distributed across enough parties that no single actor is clearly the one conducting the surveillance.

This architecture is not unique to Flock — it is endemic to modern commercial surveillance. Data broker networks, fusion centers, and joint task forces are all designed, at least in part, to frustrate accountability by ensuring that no single government actor can be said to have conducted the surveillance triggering constitutional scrutiny.

The Fourth Amendment was written for a world where a search was a discrete act by a specific officer with a specific target. The modern surveillance state is a distributed system that produces searchable records of everyone, continuously, without any individual actor ever deciding to search any particular person. Doctrine built for the former is poorly suited to regulate the latter.

What Needs to Happen

The Supreme Court needs to take a case that forces a direct answer on ALPRs and Carpenter’s reach. Lower courts need to treat the logic of comprehensive location tracking — not the specific technology — as the trigger for Fourth Amendment protection. Congress needs to pass a warrant requirement for ALPR queries, with meaningful restrictions on federal access to locally generated data.

And cities need to have the conversations they skipped. Surveillance infrastructure should require affirmative community approval, public data-sharing audits, and explicit prohibitions on feeding local data to immigration enforcement. The camera on your street corner was a policy choice. It should be treated as one.

The people climbing poles with bolt cutters are not offering a constitutional argument. But their frustration is a constitutional indictment: of a legal system that has allowed the surveillance state to outpace every check designed to contain it.

FixTheSupremeCourt.org covers the intersection of constitutional law, judicial accountability, and civil liberties. This post is part of our ongoing coverage of digital privacy and Fourth Amendment rights.

Sources