Flock Cameras and Texas Traffic Stops

This is not science fiction. It is the everyday reality of automated license-plate readers, and the dominant name in them is Flock. In a New York Times investigation published in August 2026, reporters documented how thoroughly these cameras now blanket American roads and the bipartisan backlash that has followed. Since then, the question has stopped being theoretical in Texas. The governor has cut off state money for the cameras, cities have unplugged hundreds of them, a Lufkin officer is facing roughly a hundred criminal counts over how he used the data, and in late September a federal jury in San Antonio found that a Texas county had been using plate readers to manufacture traffic stops.

For a Texas driver the practical question is unchanged and more urgent: when a camera is the reason you got stopped, what does that mean for your case, and how do you fight it?

A Federal Jury Just Called One of These Stops Unconstitutional

In March 2022, Alek Schott was driving home to Houston on Interstate 35 after a work trip to Carrizo Springs. A Bexar County sheriff’s deputy, Joel Babb, pulled him over and told him he had been “drifting over that fog line pretty hard.” Schott’s own dashcam showed that he never left his lane. He was held on the side of the highway for 76 minutes while deputies searched his truck. They found nothing.

What came out in the litigation was how the stop actually started. Before Babb ever laid eyes on the truck, license-plate readers had tracked it heading south toward the border and back again, and a law-enforcement group chat had circulated that travel pattern as suspicious. The reader hit was not a hotlist match for a stolen vehicle or an outstanding warrant. There was nothing to match. It was a profile. To stop the truck, Babb still needed a traffic violation, and the jury concluded that he supplied one that had not happened.

In late September 2026, a federal jury in San Antonio, sitting before United States District Judge Orlando Garcia, found Bexar County liable. The award was $76, one dollar for each minute Schott spent detained. The number is the headline and the least important part of the verdict. To reach the county at all, the jury had to find more than one deputy behaving badly; under the policy-or-custom standard that governs civil rights claims against a local government, it had to find that the stop came out of how the county ran its plate-reader program. It did. Schott was represented without charge by the Institute for Justice, which is now asking Judge Garcia for an injunction. The county has indicated it may appeal.

Two honest cautions: It is a civil verdict, not a suppression ruling. No court in it held that reading a plate is a Fourth Amendment search, and nothing in it binds a criminal judge deciding a motion to suppress. The system the deputies were using was Vigilant, not Flock. But the mechanism is the one that matters in a criminal case, because it is the mechanism behind a great many camera-driven stops: the camera generated a suspicion, the suspicion was not a violation of anything, and so a human being had to produce a violation to make the stop lawful. A jury has now looked at that sequence and called it unconstitutional.

What Flock Cameras Actually Are

Flock cameras are automated license-plate readers, or ALPRs: cameras that photograph every passing vehicle and use software to read the plate. Flock Safety, the dominant vendor, operated a network of roughly 120,000 cameras across 49 states as of mid-2026 and contracts with something like 40 percent of American police departments. In Texas, DeFlock, an anti-surveillance group that maps camera locations by verified crowdsourcing, counted roughly 13,000 of them statewide as of late August 2026. There is no official count, which is itself part of the problem. Not every reader is a Flock camera, but Flock is by far the most common one a Texas driver will pass. Each camera captures more than a number. It records what the company calls a vehicle signature: the make, model, and color of the car, along with distinctive features like bumper stickers, decals, and roof racks. All of it is timestamped, geolocated, and uploaded into databases that agencies can search and share with one another across jurisdictions.

Two capabilities matter for a criminal case. The first is the real-time hotlist: if a passing plate matches an entry for a stolen vehicle, an outstanding warrant, an expired registration, or an Amber or Silver alert, the system pushes an alert to nearby officers. The second is the historical database: a searchable record of where a given plate has been, and when, built from every camera it has ever passed. The reach of that system has driven a striking backlash. Documented misuse, cities canceling their contracts, and civil-liberties litigation from across the political spectrum have all followed, and in August 2026 Flock announced changes, including cutting its default data-retention period from thirty days to seven, that critics called a step forward but far from enough.

What Changed in Texas This Fall

If you read about Flock cameras in the spring and have not looked since, the Texas landscape has moved under you. On August 28, 2026, Governor Abbott ordered state agencies to stop spending state money on Flock cameras. He pointed at misuse: a Lufkin police officer facing roughly a hundred criminal counts for allegedly using Flock data to surveil eleven people over more than a year, and officers at a minimum of six Texas police departments and sheriff’s offices placed on leave or charged. The Department of Public Safety paused new installations, and Abbott separately directed TxDOT to stop putting readers along state highways.

The local fallout was fast. The Dallas Police Department deactivated 321 cameras that had been paid for with the affected state grants, while keeping more than 300 others funded by the city. Statewide, at least fourteen cities and counties have unplugged more than 900 cameras, with Plano, Leander, El Paso, Kendall County, Pflugerville, and Hutto ending contracts outright.

None of which means the cameras are gone. DPS is keeping its own network of more than 900 readers, Houston and Dallas both kept their city-funded cameras, and most of the roughly 13,000 in Texas were never state-funded in the first place. What actually changed is the atmosphere. Every one of those stories is now a document request waiting to be made, and a judge hearing a suppression motion in late 2026 has read the same headlines you have.

How a Camera Becomes a Traffic Stop

For most people, a Flock camera shows up in their life as a traffic stop. An officer receives a hotlist alert that a passing plate is associated with, say, a stolen car or a warrant, locates the vehicle, and pulls it over. From there an ordinary DWI or drug investigation unfolds: the officer approaches, claims to smell alcohol or marijuana, asks questions, and the stop that began with a camera becomes a criminal case. In that sequence, the camera hit is doing critical legal work. It is supplying the reasonable suspicion, the specific and articulable basis that the Fourth Amendment and Texas law require before an officer can detain you at all. And that is exactly where these stops are vulnerable, because an ALPR hit is only as trustworthy as three things: the read, the data behind it, and the officer’s account of what he saw you do.

Where These Stops Go Wrong

The first failure is a misread. A Flock camera runs character-recognition software, and software makes mistakes: it confuses a zero for the letter O, an 8 for a B, one state’s plate design for another’s, or a dirty or partly obscured plate for something it is not. When the software misreads, it can generate a hotlist match to an entirely different vehicle, and an officer who trusts the alert can pull over a car that never did anything wrong. These are not hypothetical errors. Around the country they have produced frightening wrongful stops, including drivers ordered out of their cars at gunpoint over a plate the system simply read wrong.

The second failure is stale or erroneous data. A hotlist is only as current as the records feeding it. The stolen car was recovered last week, the warrant was cleared, the registration was renewed, or the plate was reported on the wrong vehicle, but the entry driving the alert was never updated. The camera reads the plate correctly and still fires an alert for a problem that no longer exists, and the officer stops a driver over a status that, in reality, is not true.

The third failure is the one the Schott verdict exposed, and it is the least discussed. Sometimes the camera works perfectly and the data is accurate, and there is still no violation to stop you for. A reader can show that a car drove south and came back, or passed a particular intersection at a particular hour, and an officer can find that pattern suspicious. Suspicion about a travel pattern is not reasonable suspicion of a crime, and it is not a traffic violation. So the officer needs one, and the temptation is to find a drift over the fog line, a slow roll at a stop sign, a license-plate lamp that may or may not have been out. If the real reason for the stop was the camera and the stated reason was a violation the video does not show, the stop is unlawful no matter how well the camera performed. That is why the dashcam mattered more than the reader in Schott’s case, and why it may matter more in yours.

The Rule That Protects You: An Alert Is Not Enough

Here is the principle that turns those failures into a defense. A traffic stop is a seizure, and under both the Fourth Amendment and Article I, Section 9 of the Texas Constitution, an officer needs reasonable suspicion of an actual violation to make one. Running your plate is not itself a search, because your plate is exposed to public view, and in the Fifth Circuit that question is now closed. Porter settled it. But the stop that follows still has to be justified, and a raw computer alert, standing alone, does not justify it.

You do not have to take a defense lawyer’s word for that, because Texas police departments say it themselves. The Austin Police Department’s own written policy on license-plate readers states plainly that an ALPR alert alone does not create reasonable suspicion to justify a stop, and that before detaining anyone an officer must visually confirm that the plate actually matches the one in the alert and verify the hit against the state and national crime databases. That policy exists precisely because the department knows the readers misfire on data-entry errors, swapped plates, and misreads. When an officer skips those steps and stops a driver on the bare alert, the stop violates both the constitutional standard and the department’s own rules, and everything the stop produced, the DWI, the drugs, the statements, becomes vulnerable to suppression.

The Texas Advantage: A Narrower Good-Faith Exception

This is where Texas law gives defendants a real edge, and it is worth understanding. In federal court, when a stop rests on an erroneous database entry, the government often escapes suppression by invoking the good-faith exception, the idea that the officer reasonably relied on the system and should not be penalized for someone else’s recordkeeping error. Texas is different. Its exclusionary rule is a statute, Article 38.23 of the Code of Criminal Procedure, and it is broader than the federal rule. The only good-faith exception Texas recognizes is for an officer acting in reliance on a warrant issued by a magistrate. A warrantless ALPR stop involves no warrant at all, so that exception does not apply. In plain terms, a bad camera hit that a federal court might forgive can still get the evidence thrown out in a Texas courtroom.

What the Fifth Circuit Has Decided, and What It Has Not

Any Texas lawyer arguing about plate readers now has to start with United States v. Porter, No. 25-60163 (5th Cir. Mar. 17, 2026). It is published, it is binding in Texas, and it went the government’s way. Porter held that the government’s use of a license-plate reader system was not a Fourth Amendment search at all. The reasoning leaned on United States v. Knotts, the beeper case, and on a Fifth Circuit rule that “[a] motorist has no privacy interest in their license plate number.” The reader system, the court said, “is not capable of tracking the ‘whole of [an individual’s] physical movements,’ much less ‘for a very long period,’” because it does not “‘faithfully follow[]’ individuals ‘beyond public thoroughfares.’”

It matters that the court went further than that. Porter did not leave the accumulation question untouched for a future case. It reached the argument that a stored history of reader hits becomes something constitutionally different from a single hit, and it rejected it. The court granted that reader data lets police “travel back in time to retrace a person’s whereabouts,” then held that “the LPR technology in the instant case provides only periodic information about a vehicle’s location when a vehicle passes one of its ten locations where an LPR camera is stationed in Gautier and is much more limited than CSLI and geofence data.” Anyone who tells you Porter reserved this question has not read it.

The argument survives anyway, and the honest version of it is stronger than the version that pretends the court looked away. Every reason Porter gave is a matter of degree, and the court tied its holding to its own record with unusual care: “the LPR technology in the instant case,” one of “its ten locations” in Gautier, Mississippi. Ten cameras in one small Gulf Coast town is not thirteen thousand cameras across Texas feeding a database that hundreds of agencies can query. Porter rejected an accumulation argument on a thin record and in deliberately fact-bound terms. That leaves the Texas-scale version of the argument open, and it needs to be made on a developed record about how many cameras covered the relevant roads, how long the data was kept, and who could search it.

The Bigger Fight: Warrantless Tracking

Beyond the individual stop lies the question Porter answered narrowly and no court has answered at scale. A single hotlist hit is one thing. A historical database that can reconstruct everywhere a car has been over weeks or months is another. In Carpenter v. United States, 585 U.S. 296 (2018), the Supreme Court held that using technology to compile a detailed record of a person’s movements over time can be a Fourth Amendment search requiring a warrant, even though every individual data point was gathered in public. Carpenter involved 127 days of cell-site records from one carrier and two days from another. The whole fight over plate readers is whether enough cameras, held long enough, do the same work.

Texas has its own version of Carpenter, and it is worth knowing precisely. In Holder v. State, 595 S.W.3d 691 (Tex. Crim. App. 2020), the Court of Criminal Appeals held that the third-party doctrine “alone cannot defeat a person’s expectation of privacy in at least 23 days of historical” cell-site location information under Article I, Section 9 of the Texas Constitution. Read the limiters: “alone,” and “at least 23 days.” And do not oversell it as Texas going its own way. The court expressly refused to read the state provision differently from the Fourth Amendment, saying that to do so “would, at best, amount to us being different just because we can.” Its stated rationale is Carpenter’s. The same opinion notes that “there is no implied warrant requirement in Article I, Section 9.” Holder is binding Texas precedent and a federal retrenchment would not automatically undo it, but a prosecutor can quote that language straight back at you, and will.

Outside Texas, the courts to reach the plate-reader question have so far said no search, while flagging where the line might be. In Commonwealth v. McCarthy, 484 Mass. 493 (2020), the Massachusetts high court held that limited ALPR use, four cameras at two bridge locations, was not a search. In dictum it said a sufficiently pervasive network would be, then disclaimed the job of drawing the line: “we need not, and indeed cannot, determine how pervasive a system of ALPRs would have to be.” In Robinson v. Commonwealth, Record No. 1912-24-1 (Va. Ct. App. Apr. 7, 2026), a published decision on a record of 172 Flock cameras in Norfolk with 30-day retention, the court again found no search, and this time left the door visibly open: “We decline to speculate as to when—or if—the Flock cameras could create such ‘a comprehensive chronicle’ of a person’s movements where that person would then have a reasonable expectation of privacy.”

That same Norfolk network is the subject of the appeal to watch. Schmidt v. City of Norfolk, No. 26-1227, is fully briefed in the Fourth Circuit and squarely presents whether a 172-camera citywide system triggers Carpenter. It was brought by the Institute for Justice, the same firm that tried Alek Schott’s case. Sixteen states and the District of Columbia have filed in support of Norfolk; privacy organizations have filed on the other side. The Supreme Court’s own most recent word on digital location data, Chatrie v. United States (June 2026), is encouraging for defendants on the threshold question but settled nothing on remedy: Chatrie himself has not won suppression, and warrant particularity, probable cause, and good faith are all still open in his case.

So the state of play is honest and unglamorous. Every appellate court to decide whether plate-reader data is a search has said it is not, on the records in front of it. None has decided it on a thirteen-thousand-camera record. If the State placed your car somewhere using a historical database query rather than a live hotlist hit, the warrant argument is not a winner today, and it is the argument you have to preserve anyway, because it is the one the courts have told us they have not yet answered.

The Licensing Argument, and Why It Is Weaker Than It Looks

There is an argument circulating in Texas defense circles that deserves a careful answer rather than an enthusiastic one. Flock operated cameras in Texas for years without the private-security license state law requires. DPS issued a cease-and-desist, Flock got licensed, and in March 2026 DPS closed its investigation with a $500 fine, the maximum for a first-time offender. The argument is that evidence from cameras operated illegally was “obtained… in violation of” Texas law and must be excluded under Article 38.23.

It is a real argument and it is untested, and you should know the two cases a prosecutor will open with. In Gonzales v. State, 67 S.W.3d 910, 912 (Tex. Crim. App. 2002), the Court of Criminal Appeals held that “evidence is not ‘obtained… in violation’ of a provision of law if there is no causal connection between the illegal conduct and the acquisition of the evidence.” Chavez v. State, 9 S.W.3d 817 (Tex. Crim. App. 2000), rejects simple but-for causation and asks instead whether an ordinary person would consider the evidence to have been obtained by that illegality. And Schultz v. State, 457 S.W.3d 94 (Tex. App.—Houston [1st Dist.] 2014), says it plainly: “noncompliance with administrative agency rules does not provide a basis for the exclusion of evidence under article 38.23.”

The problem is simple: The camera photographed the same plate whether or not the vendor’s paperwork was in order, and a $500 fine makes the violation look as technical as a violation can look. No Texas court has decided whether a vendor’s licensing failure triggers Article 38.23. Make the argument if the facts invite it, make it knowing Gonzales and Chavez are sitting there, and do not build a motion on it alone.

There is a far better Article 38.23 route: The statute reaches evidence obtained in violation of law, and the Texas misuse scandals are about illegality in the obtaining. A Lufkin officer is facing roughly a hundred counts over using Flock data to track eleven people. Officers in at least six Texas agencies have been investigated, suspended, or charged over how they queried these systems. If the query that produced evidence in your case was made for a reason the system’s own rules did not permit, that is not a paperwork defect. That is the thing Article 38.23 was written to reach, and it is worth finding out with an audit-log request before you settle for a licensing theory.

If a Camera Was Behind Your Stop

This is general education, not advice about any specific case. But if an ALPR alert is anywhere in the story of how you were stopped, these are the things that matter:

Find out what actually triggered the stop. Demand the Flock alert (or other ALPR alert), the hotlist entry it was based on, and the dispatch or real-time-crime-center records behind it.
Ask whether the officer verified. Did the officer visually confirm the plate and check the hit against the crime databases before stopping you, or act on the raw alert? Departmental policy usually requires the former.
Attack the read. Was it a misread, a wrong-state match, or an obscured plate? A hit on the wrong vehicle is no reasonable suspicion at all.
Attack the data. Was the hotlist entry stale, already cleared, or simply wrong when the alert fired?
Get your own video. Compare the officer’s stated reason for the stop against every camera that saw it: his dashcam, his body camera, your dashcam, nearby business video. Alek Schott won because his own camera contradicted a sworn account of a fog-line drift.
Ask for the audit log. Flock and comparable systems record who ran each query, when, and what reason they typed in. Ask for the audit trail for every query touching your vehicle. Given what Texas agencies have been investigated for this year, it is a question worth asking in every camera case.
Move fast to preserve the records. ALPR data can be deleted quickly, in as little as seven days by default now, so an early preservation demand and discovery request can be the difference between proving the error and losing it.
Raise the tracking question. If the State used a historical database query rather than a live alert to place your car, raise the Carpenter warrant argument and build the record for it. It is not a winner in the Fifth Circuit today, and it is the question the courts have said they have not yet reached at scale.

The Bottom Line

Here is where things actually stand at the end of 2026. Every appellate court to rule has said that reading your plate is not a search, and in the Fifth Circuit that is settled law. Texas has cut state funding, cities have unplugged hundreds of cameras, and the large majority of the roughly thirteen thousand in Texas are still watching. None of that decides a single case.

What decides cases is narrower. A camera hit is a lead, not proof of a violation. It can be a misread. It can rest on data that went stale months ago. And as a federal jury in San Antonio found, it can be accurate and still not describe a crime, leaving an officer to supply a violation that did not happen. In every one of those situations the officer still had to have a real, verified basis to pull you over, and Texas, with an exclusionary rule broader than the federal one and a good-faith exception that reaches only warrants, gives you a genuine path to suppress what the stop produced. The lesson of this series holds here too: the output of a machine is only as good as the data and the human judgment behind it, and you are entitled to demand a look at all three.

Deandra Grant is the managing partner of Deandra Grant Law, a Texas DWI and criminal defense firm with offices across Dallas-Fort Worth and Central Texas. She and her team challenge unlawful stops, searches, and forensic evidence throughout the region.

Flock Cameras and Your Texas Traffic Stop