The Pender Sheriff Answered. Here’s His Statement — and What the Records Show.

On August 6, after weeks of resident questions, Pender County Sheriff Alan W. Cutler released a statement on the county’s Flock cameras. Answering the public is the right instinct, and we said so. We also do what we always do: publish the statement in full, and hold it against the records the county itself produced.

Most of the statement is accurate. Two claims are contestable, one central claim is asserted rather than shown, and the whole of it describes a system smaller than the one the county’s own invoices say it bought — and it answers almost none of the questions that decide whether Pender should be running it at all. Here is the statement, then the record, then those questions.

The statement, in full

What holds up

Several of the assurances are accurate, and we say so plainly. Flock’s ALPR product does not use facial recognition, does not identify individuals, and does not collect biometrics, names, home addresses, or phone numbers — independent reporting has not contradicted that for the plate-reader product. The statement is also right that Flock, not the Sheriff’s Office, holds the property license agreements: the landowner grants Flock the license, and Flock owns the hardware. And it is precise that the cameras sit “on both public and private property” — our own review of the sites found one on county property and the rest on private land. Credit where it is due.

One clarification, because the wording carries a lot of weight. The facial-recognition denial is accurate — and it helps to know exactly what it covers. “Facial recognition,” narrowly, means matching a face to an identity, and Flock says it does not do that. But Flock’s video product is a different matter. Its PTZ Condor camera uses AI to detect and follow a person on foot (Flock’s own training videos, surfaced by InvestigateTV in June 2026, show it acquiring and tracking pedestrians), and its FreeForm search lets an officer type a description of a person — clothing, accessories, even, per 404 Media, race — and pull matches across networked video feeds. That is person-tracking that stops just short of “facial recognition” as the term is narrowly defined. Here is the part that matters, and it is in Pender’s favor: the county did not buy any of it. Its invoices show plate-reader cameras and the FlockOS platform — no Condor video — and the person-search tools run on the video feeds the plate-readers do not provide. So the Sheriff’s denial is true, and Pender’s system genuinely cannot follow people on foot. We note it only so residents understand that “Flock does not use facial recognition” is a narrower assurance than it sounds — and that the same company sells the part that does watch people, should a larger system ever be proposed here.

The invoices describe a bigger system than the statement does

Here is the heart of it. The statement says the cameras capture “observable vehicle characteristics such as make, model, color, and vehicle type,” used “solely as an investigative tool.” But the county’s own Flock invoices — which we obtained and published — show Pender did not simply lease cameras. It licensed a software platform:

Source: Flock invoices INV-56056 (Jan 10, 2025) and INV-73504 (Sept 15, 2025), produced by Pender County. The county’s own purchase order (#451, $70,600) describes the entire package, simply, as “Flock Cameras.”

FlockOS Advanced is not a camera. It is Flock’s operating system — a real-time crime-center platform that, in Flock’s own words, unifies license-plate readers, video, drones, and other sensors on a single live map. Its capabilities run well past the four attributes the statement lists. Per Flock’s own product materials:

• Vehicle Fingerprint lets a user search not only by plate, make, model, and color, but by bumper stickers, decals, roof racks, dents and damage, wheel type, and missing or covered plates. Neighboring Greenville, N.C. lists exactly these — “decals, roof racks, or bumper stickers” — in its own public Flock FAQ.
• Advanced Search adds Convoy analysis — enter one plate, and the system returns the vehicles that habitually travel with it, a machine-built list of associates — and Multi-Geo search, which finds a vehicle across multiple locations at once. (Government Technology.)

In plain terms: with the platform Pender licensed, an officer can enter your plate and pull up every place a Flock camera photographed your car — each sighting time-stamped, GPS-tagged, and dropped on a map — as far back as the data is kept. Flock’s default, and Pender’s own written policy, is 30 days (state law allows up to 90). Run a plate through the network and what comes back is enough repeated observations to reveal a driver’s recurring patterns — the roads regularly taken, the places returned to day after day. Flock’s platform also offers Convoy analysis — give it one plate and it surfaces the vehicles that repeatedly travel alongside it. That is what this class of software is built to do. Whether Pender’s specific license and portal turn on every piece of it — Convoy and cross-network search can be tiered add-ons or permission-gated, and an invoice line does not settle it — is a separate question, and one its promised transparency portal should answer.

We are careful about what this does and does not prove. Licensing a capability is not the same as using it, and an invoice is not a feature list: it establishes what was purchased, not which software modules are switched on in Pender’s account. Flock sells the platform in editions, with tools gated by license, permission, and agency settings, so we cannot say from an invoice line which of them Pender can actually reach. Pender’s own Organization Audit — the log of its deputies’ 364 searches in a month — shows only ordinary plate lookups and searches; no Convoy or Multi-Geo entries appear in the window we have. So we are not claiming Pender deputies mapped anyone’s traveling companions. We are claiming something narrower and undisputed: the county paid for a platform that can, and the Sheriff’s four-attribute description is the smallest part of what was licensed. Residents deciding whether they are comfortable with this deserve the whole of it, not the least of it.

“Solely an investigative tool” — except the reason codes say otherwise

The load-bearing sentence is that the system is used “solely as an investigative tool” for solving crimes, stolen vehicles, and missing persons. Pender’s own search audit sits awkwardly beside it. These are not automated camera captures or hotlist alerts — they are deliberate lookups a deputy typed into the system, each with a reason chosen from a menu. In a single month, deputies ran 364 of them, and roughly one in eight carried the reason code “Traffic Infraction” — among the most common reasons in the log. North Carolina law limits ALPR use to a short list of purposes and says plainly that the data “shall not be used for the enforcement of traffic violations” (N.C. Gen. Stat. § 20-183.31(b)). The statement never mentions reason codes at all.

There are at least two ways to read that, and the innocent one deserves to go first. A deputy who stops a car for a broken taillight and then runs its plate to see whether it is stolen has made an investigative use, even though a traffic infraction is why he was standing there; on that reading, “Traffic Infraction” describes the encounter, not an attempt to write tickets by camera, and no line of the statute is crossed. Fair enough. But look where it leaves the record: the one column a reviewer would scan to catch misuse — the stated reason — is being filled with the context of a stop rather than the purpose of the search. So the audit resolves into a dilemma, and both horns are a problem. Either the searches are what they say they are, and a use the statute brackets off is among the most common in the log; or they are miscoded, and the log that is supposed to make this system accountable cannot be read at face value. We call no single search unlawful. We say only that both roads run somewhere the Sheriff’s one word — “solely” — does not.

And we will say this now, before it is said to us. When New Hanover County was shown the same pattern, its Sheriff’s Office told WECT the “traffic infraction” codes had been mislabeled. If Pender offers that same explanation, it does not close the question — it opens a larger one. The reason code is the agency’s own entry in its own audit, and that audit is the safeguard every sheriff points to when asked how misuse would ever be caught. A department cannot hold the audit log up as its accountability mechanism and, in the same breath, tell the public the log’s entries are not to be believed. If the labels are wrong, the honest fix is to correct them and show the corrected log — not to wave the number away. We wrote about exactly this problem here.

What the statement never mentions: who else is searching

The largest privacy question about Flock is sharing — who outside Pender County can search these cameras. The statement does not address it at all. The records do: in Pender’s audit, a single search reached across 6,234 separate camera networks at once, and most searches touched more than a hundred. That is not a local system that occasionally cooperates; it is a node on a national one. Here is what that network looks like across North Carolina.

The speed trailer: one integrated product, two claimed data paths

The statement says the mobile unit’s speed display “operates independently of the Flock Safety system,” and that speed data is not sent to or stored by Flock. That is very likely how it works: on these units the radar typically drives a local LED display to calm traffic, while the plate-reader takes its own stills for hotlist matches, and the two need not touch. We note it only because the invoice lists the trailer as one integrated product — the “LPR Trailer — Radar Speed and Message Sign,” a single $13,000-a-year Flock subscription — so “independent” is a statement about how one device is configured, not about two separate machines. It is an easy thing to put to rest, and in the Sheriff’s favor if he is right: the Flock configuration for that unit would show whether the plate capture and the radar reading ever share a record.

On the pilot timing, the Sheriff has a point

The statement says that when Pender signed on, “the pilot program jointly through NCDOT and the SBI was not available.” The timing is close, and it runs the Sheriff’s way. Pender authorized its Flock contract on December 19, 2024. By the SBI’s own account, it did not identify the pilot’s initial participating agencies until January 2025; and Pender’s SBI MOU carries a February 2025 effective date (the day left blank; the copy in hand was physically signed in April 2026). Every one of those dates falls after the county had already committed to Flock. So the Sheriff has a fair point: when Pender bought, the state right-of-way route was not yet in place for it.

That February 2025 effective date also answers the obvious question — if the MOU was not signed until April 2026, how did the SBI’s April 7, 2026 report already count Pender among its 32 participating agencies? Because participation runs from the MOU’s effective date, not the day the ink dried — the county was on the roster before the signatures caught up. What none of this explains is why the cameras still sit on private and county land today, now that Pender is enrolled and the right-of-way route exists, where an encroachment permit would apply. But that is a small question next to the ones the consent-agenda vote and the nationwide sharing raise, and none of those turn on it.

The questions a statement cannot settle

Everything above is about what the system is. The harder questions are whether Pender County should be running it at all — and the statement does not reach them. Here they are, held against the same record.

Did anyone actually vote for this? Yes — and here Pender differs from New Hanover County, whose Sheriff bought in on a contract the elected Board never voted on. Pender’s commissioners did vote, and the county’s own minutes show it. On December 19, 2024 — the organizational meeting where newly elected commissioners took the oath of office — Sheriff Alan Cutler asked the Board, in a written memo, to authorize a Flock Safety contract. The memo described the product in a single sentence: “Flock Cameras… which are used around the nation to help minimize crime rate and solve crimes.” It put the county’s share at $57,600, with another $13,000 from the “Sheriff’s Charitable Foundation” covering the rest of the $70,600. The Board authorized it. Then, on January 21, 2025, it funded it — as consent-agenda item 8.6, approved in a single block alongside a turf planer, mosquito-spray chemicals, a vaccine order, and the county’s 150th-anniversary costs. All five commissioners were present: Brad George, Jerry Groves, Chairman Randy Burton, Brent Springer, and Ken Smith. During public comment that night, residents spoke about ballfield lighting and the county anniversary. No one spoke about Flock — it was a consent-agenda line, not an announced decision.

So this was not slipped in behind the public’s back; it was approved on the record. Which makes the real question the next one: what was the Board told it was buying? The county’s decision memo described it as “Flock Cameras… used around the nation to… solve crimes.” The slide deck the Sheriff presented went further, on a slide headed “Privacy Commitment.” It assured the Board that Flock devices “do not connect to databases to identify… people,” that every search “requires a documented search reason or case number… creating a permanent audit trail,” and, word for word, that “Customers own 100% of all data captured… We never share or sell data with third parties.”

Hold those assurances against the county’s own records. “We never share… with third parties” is a line a commissioner hears as your data stays here — yet a single Pender search reaches 6,234 outside networks, and police in other states and the FBI search Pender’s cameras. Flock’s defense is a definition: sharing among law-enforcement agencies on its own network is not, in its telling, sharing with “third parties.” Perhaps — but that is not what the words convey, and it is not what the Board was invited to weigh. The “documented search reason” and “permanent audit trail” were offered as the safeguard — the same audit whose reason field, in Pender’s own log, is filled with the context of a stop rather than the purpose of the search. A vote to buy “cameras that solve crimes,” sold with a promise that the data never leaves, is not the same as a decision to join a national surveillance grid. On this record, the commissioners cast the first; the second was never put before them in terms they could recognize. (Primary sources: the December 19, 2024 special meeting and the January 21, 2025 meeting.)

One more thing, on accountability. The board that approved Flock has largely turned over since. Of the five who cast that vote, four are leaving office: Chairman Randy Burton lost his 2026 Republican primary; Commissioners Brad George and Jerry Groves did not file for re-election; and Commissioner Ken Smith resigned after relocating out of the county. Only Brent Springer remains, his term running through 2028. So the decision to make Pender a node on a national surveillance network was made, in large part, by a board now on its way out the door — and none of the five is on the 2026 ballot to answer for it.

Do Pender’s residents want it? The meetings where it was approved were public and carried a comment period, so the door was technically open. But the item moved as a line on a long agenda under the plain label “Flock Camera System,” and there is no sign the public grasped that a nationwide, out-of-state-searchable network was what was being bought. The “rising community questions” the Sheriff is now answering are the first time residents have engaged it in earnest — and only after the cameras were already watching.

Did the Sheriff enroll Pender in a national surveillance network — the way New Hanover’s Sheriff did? The records say yes. A single Pender search fanned out across 6,234 outside camera networks; it is the same shared system that lets police in Houston, the Texas Department of Public Safety, and the FBI search North Carolina’s cameras. Pender’s plate data does not stay in Pender — and the states it can travel to, Texas and Florida among them, have no ALPR statute barring its sale or open sharing. North Carolina’s protections stop at the state line; the data does not.

What does the evidence say it buys? Less than the price implies. The peer-reviewed record on plate readers is thin and mixed: rigorous evaluations are few, the clearest benefit is narrow (recovering stolen vehicles), and a 2025 evaluation of a major ALPR expansion found it did not reduce violent crime (Justice Evaluation Journal). Independent research showing that blanketing a county in readers is a cost-effective use of tax dollars is nearly absent. Pender is paying roughly $76,000 over two years for a tool whose general effectiveness the research does not establish.

Is anyone independent checking it — and will the public hear about abuse? By statute the only required audit is annual, internal, and self-conducted (N.C. Gen. Stat. § 20-183.31): the same office that runs the system reviews its own use of it. State law requires no independent outside audit, and no security or network audit at all. Pender has promised a “transparency portal,” which is welcome — but it has not said whether that portal will disclose misuse when it happens, or whether residents will learn of it the way Charlotte and Mooresville did: only after an officer was criminally charged.

The bottom line

This is a more transparent response than most agencies give, and the promised portal is welcome. But a statement is measured by what it includes. This one answers the small questions — biometrics, license holders — and leaves the large one untouched: should a free county log the movements of tens of thousands of people suspected of nothing, on the chance that a few are worth finding? No product specification answers that. Pender’s Board did vote to buy the system — but a vote to approve “a Flock camera system” is not the same as a public reckoning with what that system is: a node on a nationwide network, searched from other states, feeding data to places North Carolina’s law cannot follow. That reckoning has not happened. It still can — in public, before the Board that approved it and can end it. The county’s own records — which we have published in full so anyone can check us, and the Sheriff, against the same paper — are where it should start.

Sheriff Cutler’s statement is reproduced in full as a public record, as provided to WECT (Aug 6, 2026). Product capabilities are drawn from Flock Safety’s own published materials and Government Technology’s reporting; we describe capabilities the county licensed and do not allege that any specific search was unlawful or that any advanced feature was used. All Pender figures are DeFlockILM’s tally of the invoices, purchase order, and search audits Pender County produced under a public-records request, archived in full at the Internet Archive. This article is general information and commentary, not legal advice, and creates no attorney-client relationship. If any figure here is wrong, write to mark@deflockilm.org and we will correct it in public. Published August 6, 2026.