SURVEILLANCE / INVESTIGATIONS — NOVEMBER 4, 2026. Grace Knowles says she was driving toward Florida when an officer pulled her over and eventually showed her a digital map of where her car had traveled. He was not showing her evidence of speeding, swerving, or running a red light. According to her account, the troubling thing was the route itself: a computer had marked her travel history as suspicious.
The encounter happened in September 2025, but Knowles first told her story publicly in an October 26, 2026 report by the American Civil Liberties Union. Her experience sits at the intersection of three developments: license-plate databases that record cars passing fixed points, analytics that look for unusual movement patterns, and police stops justified by ordinary traffic infractions. The result is a question with real Fourth Amendment stakes: Can where you have driven become the reason police decide to stop you?
A road trip, a whiskey-museum pamphlet, and a map of her movements
Knowles, then a woman in her 20s from northern Virginia, told the ACLU that her road trip took her through Oregon, Texas and Louisiana. After leaving New Orleans on her way to Florida, she noticed a police vehicle following her. The officer said she had remained in the left lane too long, she recalled. She disputed the implication that she had been driving improperly.
What came next was stranger than the alleged lane violation. Knowles said the officer asked where she had been and pressed her to prove she had visited New Orleans. She produced a pamphlet from a whiskey museum. Only after that, she told the ACLU, did he explain that he was connected to the FBI and that her license plate had returned a questionable travel history. He reportedly suspected she was lying because the system had no recorded plate sightings in New Orleans.
The officer then invited her into his vehicle to see the display. Knowles described a U.S. map with markers in places she had traveled; selecting them displayed photographs of her plate. She also said the officer described a project concerning drug or human trafficking.
Important limit: this is Knowles’s account, as relayed to the ACLU. The published account does not independently establish the officer’s identity, identify the software used, include the stop report or prove that the specific platform was Flock Safety. It would be inaccurate to present the episode as a confirmed “Flock stop.” The ACLU itself discusses Flock alongside other vendors and policing programs.
What the cameras know—and what they do not
Automated license plate readers, or ALPRs, capture images of vehicles passing certain locations. Software extracts a plate number and can store associated information: the time, location, direction and, depending on the system, vehicle make or color. This is not the same as watching a car continuously with a GPS tracker, nor does an ordinary roadside plate reader by itself determine whether the driver was speeding or weaving. Flock’s own description of its license-plate cameras emphasizes that each sighting is a snapshot rather than continuous live tracking and that its ALPR cameras do not use facial recognition.
But a snapshot repeated often enough becomes a travel history. Police can search for a specific vehicle across multiple camera locations; advanced tools can also compare journeys or highlight associations among vehicles. Absence from a camera database, crucially, does not establish absence from a city. A driver may use an uncovered route, pass behind an obstruction, have a plate missed, or travel through an area with no participating camera.
Flock did advertise automated cross-state alerts
The existence of pattern-based tools is not speculative. In a February 13, 2025 product announcement, Flock described a feature called Multi-State Insights, intended to alert law enforcement when vehicles already considered suspect appeared in multiple states over a 30-day period. The company framed it as a tool against organized retail theft and drug or human trafficking. The key word in the vendor description is suspect: the announcement does not establish that any interstate traveler automatically becomes a suspect, and it does not identify the product involved in Knowles’s encounter.
Yet the direction of the technology is unmistakable. An ACLU analysis from July 2025 raised questions about tools that generate leads from movement patterns rather than simply following an existing lead. In August 2026, WIRED examined Flock’s developing OS Investigate system, which combines plate sightings with other police and commercial records. Flock said the product was being tested with a limited group of law-enforcement partners and was still in development. Those findings should not be confused with proof that any one officer used that product.
Not just a hypothetical: the Border Patrol program
An extensive Associated Press investigation published in November 2025 documented a federal Border Patrol program that collected plate readings, used patterns to select vehicles for scrutiny and passed tips to local officers. It described so-called whisper or intel stops, in which a seemingly ordinary traffic infraction could mask the investigative origin of the stop.
AP reported that federal agencies obtained plate information from several sources—including their own equipment, other government networks and commercial services. It also reported that Flock had temporarily provided CBP pilot access to some cameras, but the company said those pilot programs were paused. The presence of Flock data in some federal arrangements does not make every Border Patrol operation a Flock operation. CBP told AP that its systems operated under federal law, constitutional protections and multiple layers of policy safeguards.
A Texas jury showed why the legal distinction matters
A separate case produced a more concrete legal finding. On September 28, 2026, a federal jury found Bexar County, Texas, liable for violating motorist Alek Schott’s Fourth Amendment rights after a 2022 surveillance-linked stop. According to Schott’s lawyers at the Institute for Justice and independent Houston Chronicle reporting, deputies claimed a lane-related infraction contradicted by dashcam evidence. Schott was detained for 76 minutes, his truck was searched, and officers found no contraband. The jury awarded him the $76 he had requested.
The verdict concerned the county’s conduct and the stop-and-search program in that case. It was not a nationwide ban on ALPRs, a holding that every AI-generated tip is unlawful, or a finding about Knowles’s encounter. Legal challenges and possible appeals can change how an individual judgment is implemented.
What does the Fourth Amendment actually require?
Police ordinarily need a lawful justification to seize a driver for a traffic stop. A brief investigative stop requires reasonable suspicion supported by specific, articulable facts; an actual traffic offense can independently justify a stop. In Whren v. United States (1996), the Supreme Court held that a stop supported by probable cause of a traffic violation is not automatically unconstitutional merely because officers also have another investigative motive. In Kansas v. Glover (2020), the Court allowed a stop based on a particular license-record inference, not an unexplained algorithmic hunch.
There is no simple nationwide ruling that any unusual driving itinerary, standing alone, is reasonable suspicion. Courts also distinguish collecting or searching plate data from stopping the person behind the wheel. The Fifth Circuit, for example, upheld a 2026 stop in United States v. Porter tied to a confirmed outstanding arrest warrant—not merely to a route labeled odd. The precise facts, search duration, jurisdiction and reason asserted for detention matter.
The problem of an algorithm that cannot explain itself
For police, networked license-plate data can provide valuable leads in searches for stolen cars, missing people and suspects linked to specific crimes. For drivers who have done nothing wrong, the same network raises uncomfortable questions. How does an analytics model distinguish a delivery driver, an erratic vacation itinerary or a family emergency from activity associated with trafficking? How often is the system wrong? Who reviews the alert? And can a stopped motorist ever see the evidence used against them?
The answer cannot be a slogan that cameras are always good or always bad. Agencies need concrete measures of effectiveness, documented access controls, independent audits, disclosed sharing agreements and enforceable limits on unnecessary stops. Flock says customer agencies control their data access and that searches are logged. Those are relevant safeguards, but the existence of a policy does not by itself establish whether every real-world search or stop complied with it.
The missing pieces of Knowles’s encounter
The public account does not include a police incident report, a body-camera recording, an agency confirmation of the officer’s assignment, or a record of the specific analytic alert. Nor does it show that Knowles was charged with any offense. Those gaps should stay visible rather than being filled with assumptions.
The supplied YouTube Short, “Suspicious Driving Patterns”, is a commentary summary of this issue, not footage of Knowles’s stop. Apep Bar has previously examined the wider system in Flock Fever and the local backlash in Standing Up to Flock. This follow-up focuses on the individual stop and the boundary between an investigative lead and lawful detention.
Commentary short about license plate surveillance; this is not footage of the reported road-trip traffic stop.
The question: If traveling through several states can make a computer suspicious—and failing to appear on its map can make an officer doubt your story—who, exactly, gets to travel without having to explain themselves?
Video and documentary sources
Watch the supplied commentary Short (0:46; not actual stop footage). Primary and major reporting: ACLU / Jay Stanley, October 26, 2026; Flock’s 2025 feature announcement; AP’s 2025 investigation; Institute for Justice / Schott verdict; Whren opinion; Glover opinion.
Featured image: Illustrative public-space surveillance photograph by Igor Savelev / Unsplash, previously credited in the Apep Bar media library. This photograph does not depict the specific Flock hardware, Grace Knowles, or the reported traffic stop. Visual documentary reference: A genuine Flock ALPR camera was photographed by Tony Webster in Jackson County, Colorado, May 25, 2026; see the photograph and CC BY 4.0 license on Wikimedia Commons.
