What Does a Flock Camera Actually Prove?
Flock Cameras: Investigative Power, Privacy, and Proof, Part 2
Understanding vehicle observations, automated plate recognition, corroboration, and the evidentiary limits of ALPR data
By R. Ryan Rider, Ph.D.
A Camera Hit Feels More Definitive Than It Is
Imagine a violent offense occurs at 10:14 p.m. Investigators develop a possible vehicle description, search a Flock Safety automated license plate reader network, and locate a matching vehicle several blocks away at 10:21 p.m. The system returns a plate number, and registration information connects that plate to a named individual.
It is easy to compress that sequence into a single statement: “Flock put the suspect near the scene.” Yet that sentence crosses several evidentiary steps. What did the camera actually record? What did the software interpret? What information came from another database? What connects the vehicle to a particular driver, and what connects that driver to the offense?
Those questions are not merely semantic. In an investigation, suppression hearing, deposition, or courtroom, the difference between observation and inference can become the difference between a defensible conclusion and an overstated one.
Flock detections can be highly valuable evidence. They can document vehicles at observation points, generate leads, connect events across time, and contribute to the reconstruction of vehicle movement. Their evidentiary strength, however, depends on distinguishing among what the camera observed, what the computer interpreted, which databases were associated with that observation, and what the totality of the evidence ultimately proves.
What a Flock Camera Actually Records
Flock Safety describes its automated license plate reader, or ALPR, system as collecting license plate images, vehicle characteristics, date and time information, and camera location information. The company also states that its ALPR system does not collect driver information, biometric data, or facial recognition data. In practical terms, the system is designed to document vehicles, not identify the people operating them.
The International Association of Chiefs of Police, or IACP, describes ALPR technology in similar terms. A camera captures an image; optical character recognition converts the visible plate information into searchable characters; and the resulting information may be compared with external databases. Identifying a registered owner requires a separate query to motor vehicle records. The camera observation and a person's identity are therefore not the same evidentiary proposition.
The underlying camera image may show a vehicle and visible characteristics. The system may generate a plate translation and classify characteristics such as color, make, model, or body type. Each element can be useful, but they are not identical forms of evidence. A vehicle image is a camera observation. A plate character string or vehicle classification is a machine-generated interpretation. A registration return comes from another data source. A conclusion about who was driving is an investigative inference that must be supported separately.
The Four Layers of Flock Evidence
A useful way to evaluate ALPR evidence is to separate it into four layers: camera observation, machine interpretation, database association, and investigative inference.
The first layer is the camera observation, the underlying image, and information directly associated with the observation point. The question is: What is objectively visible, and where and when does the system report the image was captured?
The second layer is machine interpretation. Optical character recognition and vehicle classification convert visual information into searchable data. The question becomes: What did the software interpret from what the camera saw?
The third layer is database association. Investigators may use a plate number to obtain registration information, compare it with stolen-vehicle records, receive a hotlist alert, or connect it to other law enforcement information. Those records add context but originate outside the image.
The fourth layer is investigative inference. This is where investigators ask who was driving, why the vehicle was there, whether it was connected to the offense, and what conclusions are supported when the observation is considered with the remaining evidence.
The chain becomes progressively more inferential as we move through those layers. That does not make the later layers invalid. It means investigators should remain precise about which proposition each layer supports. Asking whether “Flock is accurate” is therefore often too broad. A better forensic question is: Accurate at which layer?
This approach is consistent with current professional law enforcement guidance. In August 2026, IACP advised agencies to treat an ALPR alert as an investigative lead rather than conclusive evidence and to confirm the plate, jurisdiction, record status, and connection to the driver before enforcement action. That guidance reinforces an important principle: technology can accelerate an investigation without eliminating the investigator’s responsibility to verify what the technology reports.
The Computer Read Is Not the Photograph
Flock Safety’s current License Plate Reader Policy recognizes this distinction. The company states that although low-confidence plate reads are filtered and the system has a high accuracy rate, a plate translation can still be incomplete or inaccurate. Flock instructs users to confirm the computer translation before taking action based on an alert or search.
That matters because the camera image and the generated character string are not necessarily interchangeable. If a photograph appears to show ABC1238, but the automated translation reports ABC123B, the photograph remains the underlying visual evidence while the character string represents the software’s interpretation of that evidence.
The possibility of a recognition error does not, by itself, establish that ALPR technology is unreliable. It does establish why an automated interpretation should be verified against the underlying image before investigative or enforcement decisions depend upon it. The technology can rapidly narrow a search and identify important observations, while the investigator remains responsible for evaluating the record upon which the automated result is based.
A Flock Camera Can Identify a Vehicle, Not Necessarily the Driver
The most important evidentiary distinction in many ALPR investigations is simple: vehicle identification and driver identification are not the same thing.
A camera may document a vehicle displaying a particular plate at a specific observation point. A registration inquiry may associate that plate with an individual or business. Neither fact, standing alone, necessarily establishes who was operating the vehicle at that moment.
Vehicles may be driven by family members, employees, friends, customers, or other authorized users. Rental and fleet vehicles add additional complexity. Vehicles can be sold before databases are updated, plates can be stolen or cloned, and vehicles themselves can be stolen. The point is not that these circumstances exist in every case. Their existence demonstrates why registration establishes an association, not necessarily contemporaneous operation.
Flock’s own materials state that its ALPR system does not identify drivers or passengers. Current IACP guidance reaches the same practical conclusion by advising officers to confirm the connection between the plate information and the driver before enforcement action.
A camera may therefore document a vehicle at an observation point. Additional evidence is normally required to establish that a particular person was operating that vehicle at that time.
A Flock Alert Is Not Necessarily Proof of the Underlying Fact
Texas case law provides a useful example. In Sly v. State, a White Settlement police officer received a Flock notification concerning a silver Kia Rio reported as stolen. The officer separately confirmed through the Texas Crime Information Center and National Crime Information Center that the vehicle had been reported stolen.
On appeal, the court explained that testimony about the Flock notification was not offered to prove that the Kia was actually stolen. Instead, it explained why the officer began looking for the vehicle.
That difference matters. “The system generated an alert indicating that this vehicle was stolen” is not the same proposition as “the State has independently established that this vehicle was stolen.” Hotlist entries and agency alerts can be valuable investigative leads, but when the underlying fact matters, strong investigative practice is to verify that information through the appropriate source.
This distinction also applies beyond stolen vehicles. A plate may be associated with a wanted person, an investigative bulletin, an agency hotlist, or another record. The alert indicates that an association exists within the system. It does not necessarily establish the accuracy, currency, or ultimate significance of the underlying information.
One Camera Does Not Automatically Establish a Route
Multiple ALPR observations may show a vehicle at different places and times, creating an apparent path across a city or region. If Camera A records a vehicle at 10:02 p.m. and Camera B records the same vehicle at 10:14 p.m., investigators have two documented observation points. Those points may support an inference concerning movement, but they do not necessarily document the exact route, every stop, or what occurred between the cameras.
This is familiar territory in forensic reconstruction. Two known points can define a relationship in time and space without revealing everything that occurred between them. Additional ALPR observations, surveillance video, toll records, telematics, witnesses, cell phone data, or other evidence may strengthen the movement analysis.
The investigator’s language should reflect the evidence actually available. Saying that a vehicle was observed at two locations is different from stating that investigators have reconstructed its complete path between those locations. The latter conclusion may ultimately be justified, but it requires supporting evidence.
Corroboration Is Where Flock Evidence Becomes Stronger
Recognizing the limits of a Flock observation does not diminish its value. ALPR evidence can become especially persuasive when independent sources converge on the same conclusion.
Wills v. State illustrates that process. A Flock camera at an apartment complex captured the license plate of a car associated with the appellant entering before a murder. The record, however, also included Instagram communications arranging a meeting, text messages, cell phone location data, witness testimony, ballistic evidence, and later internet search activity. The Dallas Court of Appeals considered the cumulative circumstantial evidence sufficient to support the jury’s finding on identity.
The Flock record did not need to prove the entire case. It contributed one fact to a larger evidentiary structure. Surveillance may show who exited a vehicle, cell data may place a device in the same area, a witness may identify the driver, and physical evidence may connect a person to the offense.
The value of a Flock record often increases not because the camera says more, but because independent evidence corroborates it. That convergence allows investigators and fact finders to evaluate whether separate sources point toward a common conclusion.
Reconstructing Vehicle Movement Through Multiple Observations
Medrano v. State provides a clear example of ALPR data contributing to event reconstruction. Investigators searched the Dallas Flock system after an aggravated robbery. A camera photographed the victim’s red Toyota 4Runner as he traveled home.
Approximately seven seconds later, the same camera photographed a silver SUV matching the suspect vehicle description and captured its plate.
That sequence established a specific temporal relationship between two observations of the vehicle. Investigators then used plate information, another license plate reader system, surveillance, physical evidence, forensic testing, and cell phone records to continue developing the case.
The disciplined wording matters. The Flock observation established that the victim’s vehicle passed the camera and that a matching silver SUV passed the same observation point approximately seven seconds later. The significance of that sequence was then assessed in light of the remaining evidence.
A single image may provide only a point. Multiple documented observations can create a sequence. As in crash or crime scene reconstruction, the methodology should remain the same: document first, interpret second.
Proximity Is Evidence, but Proximity Is Not Guilt
A recent federal civil case from the Northern District of Texas provides an important counterpoint. In Tovar v. Rodriguez, the court considered, in the context of a qualified immunity and Franks challenge, an investigator’s reliance on a Flock image showing the plaintiffs’ truck approximately five blocks from a shooting about 24 minutes afterward. The individuals associated with the truck were ultimately cleared.
The court nevertheless concluded that the investigator was not unreasonable in considering the Flock image supportive of probable cause when evaluated with the temporal and geographic proximity and the other information available at the time.
The procedural context matters because Tovar should not be read as announcing a broad rule that proximity alone establishes guilt or even probable cause in every case. Instead, it illustrates a narrower and more useful principle: a vehicle observation may reasonably contribute to an investigative assessment without ultimately establishing criminal responsibility.
A fact may begin as an investigative lead, develop into evidentiary support, and eventually contribute to proof of an ultimate fact. Problems arise when those stages are collapsed into a single conclusion.
Flock Camera Evidence: Authentication and Accuracy Are Different Questions
When ALPR material moves from an investigation into court, authentication becomes another issue.
In the unpublished 2025 decision United States v. Moore, the Sixth Circuit considered a challenge to Flock images containing location, date, and time information. An investigator testified about his familiarity with the system but acknowledged that he had never independently verified the accuracy of the timestamps. The district court admitted the evidence and instructed the jurors that they remained responsible for deciding how much weight to give it and how reliable it was.
The Sixth Circuit found no abuse of discretion in the authentication ruling. The investigator had experience using the system and explained how it generated the images and associated information, while the defense had not shown that the photographs inaccurately represented the scenes depicted.
The distinction is important. Under Federal Rule of Evidence 901, authentication generally asks whether sufficient evidence supports a finding that an item is what the proponent claims it to be. Weight asks how persuasive or reliable the fact finder should consider that item once admitted.
Evidence does not become infallible merely because it is admissible. Conversely, a question about reliability does not necessarily make evidence inadmissible. Those questions can operate at different stages of the evidentiary analysis.
Preservation, Audit Trails, and the Chain of Digital Evidence
Flock’s current public materials reflect an ongoing change in retention practices. Its June 2026 License Plate Reader Policy describes a rolling seven-day default, while an August 2026 Evidence Policy still contains language describing a 30-day standard retention period. On August 13, 2026, Flock announced that its recommended or default retention period for new law enforcement customers would be reduced to 7 days, while existing customers would retain their previously approved retention periods.
The practical lesson for an investigator is therefore not to assume that a universal Flock retention period applies. The relevant questions are what retention setting was in effect for the agency, what the customer agreement required, what local policy or law required, and whether the evidence was preserved before routine deletion occurred.
Flock also states that system queries are logged for auditing. When a Flock record becomes important, investigators should therefore be able to explain more than what appears in a screenshot. They should understand who retrieved the observation, when it was retrieved, what search terms and time parameters were used, whether the underlying image was reviewed, and how the record was preserved. Search logs, audit history, metadata, exported records, and associated reports may become relevant depending on the issue being litigated.
The most useful forensic question is not simply, “How long does Flock keep data?” It is, “What was the retention, retrieval, and preservation process for this evidence in this case?”
That question shifts the analysis from a general vendor policy to the actual evidentiary history of the record being offered.
What Texas State and Federal Courts Are Beginning to Tell Us
The developing cases do not suggest that Flock evidence is inherently conclusive or inherently weak. They show that the evidentiary value depends on the proposition being proved and the surrounding evidence.
Sly demonstrates that a Flock alert may explain investigative action without independently proving the factual assertion behind the alert. Wills shows how a vehicle observation can contribute to identification when it converges with substantial independent evidence.
Medrano demonstrates how closely timed observations can help reconstruct an investigative sequence. Tovar, viewed within its specific federal civil procedural context, illustrates that temporal and geographic proximity may legitimately contribute to investigative decision-making even when the people associated with the vehicle are later cleared.
Taken together, these cases suggest that the better question is not whether Flock evidence is simply “good” or “bad.” The better questions are: What fact is this record being offered to prove? What comes directly from the image? What comes from software or another database? What requires inference? What independently corroborates that inference?
Those questions move the discussion away from enthusiasm for or skepticism about technology and back toward evidence.
Investigative Intelligence Versus Courtroom Proof
Investigative tools and courtroom proof serve different functions. An investigative tool can be extraordinarily valuable because it narrows possibilities, identifies vehicles for follow-up, reveals observation points, or helps test a timeline. None of those functions requires every search result to independently prove an ultimate legal fact.
That distinction is reflected in current IACP guidance, which advises law enforcement to treat an ALPR alert as an investigative lead rather than as conclusive evidence, and to verify the plate, record status, jurisdiction, and connection to the driver before taking enforcement action.
Before describing a Flock result, investigators should ask three questions:
What did the camera actually observe?
What information was generated or obtained after that observation?
What independent evidence supports the conclusion being drawn?
The goal is not to minimize the technology. It is to use it at its highest evidentiary value, applying its strengths for detection and reconstruction while describing its outputs with the precision expected of other forensic evidence.
Describe What the Evidence Shows, Not What You Want It to Show
Consider the statement, “Flock proved Smith was at the crime scene.”
A more precise description would be: “A Flock camera approximately three blocks from the scene recorded a vehicle bearing a plate registered to Smith at 10:21 p.m.”
If surveillance then shows Smith operating that vehicle shortly beforehand, that fact can be added. If cell phone records and witnesses independently place Smith in the area, those facts can also be added. The conclusion becomes stronger because the evidentiary chain is visible rather than compressed into a single assertion.
That is sound investigative practice. Precision allows another investigator, attorney, expert, judge, or juror to see which conclusions arise directly from evidence and which depend upon reasonable inference.
Flock Safety and other ALPR systems can substantially accelerate vehicle identification, investigative searches, and the comparison of observations across participating jurisdictions. Their speed and reach make disciplined interpretation more important, not less.
A Flock detection can be a powerful piece of evidence, but the strength of the conclusion depends on whether investigators preserve the distinction between what the camera observed, what the computer interpreted, what databases are associated with that observation, and what the totality of the evidence ultimately proves.
Precision does not weaken an investigation. Precision makes the conclusion defensible.
Looking Ahead: From Evidence to Oversight
Part 1 examined why automated license plate reader systems spread so rapidly through American law enforcement. This article has focused on what an individual detection can, and cannot, establish as evidence.
The next question is broader. A single vehicle observation may reveal relatively little. Large networks of vehicle observations, particularly when participating agencies authorize cross-jurisdictional sharing, raise a distinct set of questions about access, retention, information sharing, oversight, misuse, constitutional boundaries, and public trust.
Part 3 will examine that next stage of the Flock camera debate: when an effective investigative tool becomes a question of policy, accountability, and legitimacy.
Need assistance evaluating digital, video, scene, or other investigative evidence? Contact Triple R Investigations to discuss the evidentiary issues in your case.
References
Federal Rules of Evidence, Rule 901. (2024). Authenticating or identifying evidence. U.S. Courts. Federal Rules of Evidence
Flock Safety. (n.d.). Civil liberties & rights safeguards. Retrieved September 13, 2026. Flock Safety rights and safeguards
Flock Safety. (2026a, August 12). Flock evidence policy. Flock Evidence Policy
Flock Safety. (2026b, June 30). License plate reader policy. Flock License Plate Reader Policy
International Association of Chiefs of Police. (n.d.). Automated license plate recognition. Retrieved September 13, 2026. IACP Automated License Plate Recognition resource page
International Association of Chiefs of Police. (2026, August 20). IACP statement on the responsible use of automated license plate recognition technology. IACP responsible use statement
Langley, G. (2026, August 13). Flock updates privacy, accountability, security, and transparency safeguards. Flock Safety. Flock safeguards announcement
Medrano v. State, No. 05-24-00387-CR (Tex. App.—Dallas Oct. 24, 2025) (mem. op., not designated for publication). Medrano v. State opinion
Sly v. State, No. 02-23-00198-CR (Tex. App.—Fort Worth May 23, 2024) (mem. op., not designated for publication). Sly v. State opinion
Tovar v. Rodriguez, No. 3:23-CV-1758-K (N.D. Tex. Aug. 25, 2026). Tovar v. Rodriguez opinion
United States v. Moore, No. 24-5681 (6th Cir. Aug. 5, 2025) (unpublished). United States v. Moore opinion
Wills v. State, No. 05-23-00167-CR (Tex. App.—Dallas Oct. 25, 2024) (mem. op., not designated for publication). Wills v. State opinion
Comments