One Request. Three Excuses. One Hundred Days.

On June 30, 2026, we sent an identical public-records request to three New Hanover County beach towns — Carolina Beach, Kure Beach, and Wrightsville Beach. Same nine items, same day: any ALPR vendor contract, the policy governing the cameras, the retention period, every data-sharing agreement, the Flock Organization Audit, Network Audit, and Event Log for March–May 2026, the camera counts and locations, and the invoices.

Carolina Beach produced in about six days. Kure Beach in about ten. Both released the per-search audit that names who searched their cameras. Wrightsville Beach acknowledged the request, forwarded it to its police chief — and then referred the whole thing to a litigation law firm. One hundred days later, the one record that matters most still has not come. What has arrived instead is a moving target.

Three excuses, each abandoning the last

Here is the tell. If the per-search audits genuinely did not exist, the Town’s answer would never change. Instead it has changed every time we pin it down. Read the reasons in sequence and the problem becomes obvious: they cannot all be true, and each new one concedes that the one before it was wrong.

1. “You never requested those records.” (Aug 17) The Town’s counsel, Susan G. Renton, opened by asserting the audit items were not part of our request — in a letter that reproduced our June 30 request in full, audit items and all. We quoted the request back to her, line by line. The excuse was abandoned.
2. “They’re outside the software’s lookback period.” (Sep 1) Next the position became that the Organization and Network Audits for March–May simply weren’t reachable. But the New Hanover County Sheriff’s Office — same county, same vendor — produced its Flock audit from January 17, 2025 through June 5, 2026, more than sixteen months, which we pointed the Town to by name. If the Sheriff next door can pull sixteen months, “outside lookback” is a setting, not a wall — and the Town can direct Flock to produce it. The excuse was abandoned.
3. “The Town need not create records.” (Oct 7) Finally, the fallback that these are records the Town would have to “create.” They are not. The Organization and Network Audits are one-click exports from Flock’s own dashboard — records that already exist (we show exactly how below). And in the same breath, counsel enclosed a document the Town plainly did export on demand: an offense-type usage summary. The “can’t create it” argument arrived attached to proof that the Town can.

Three reasons, each contradicting the next. That is not what a records custodian with nothing to hide sounds like.

What the Town did produce — and what it reveals

Give the Town its due: along with the contract and policy, it produced an Event Log — the account-activity log of hotlist edits, user changes, and camera-sharing grants. That is a real record, and we credit it. But two things about it matter. First, it is not the per-search audit we are still asking for: an Event Log shows that a hotlist was edited or a network was shared, never who ran a license plate, for what reason, or on what case. It came, moreover, as a scanned PDF rather than the native CSV the records law contemplates for a database export — so even the log we got arrived stripped of its sortable fields.

Second, read it and it argues for the records the Town is withholding. In April and May 2026 alone, the log shows Wrightsville Beach granting search, alert, or video access to a parade of outside agencies — Lincolnton, Mebane, Garner, Havelock, Greenville, Ahoskie, Troutman and Newton police departments, and the Edgecombe, Granville, Caswell, Surry, Davie, Polk, Johnston, Vance, Martin, and Union county sheriffs, among others. The town’s four cameras are wired to agencies across the state. The one record that would show how often all of them actually searched those cameras is the Network Audit — the record the Town still will not produce.

The safeguard Flock built — and the Town’s log never shows switched on

There is one more thing the Event Log does not show, and it may be the most telling. In April 2026 — announced April 14 — Flock released a tool called Audit Assistance, offered free to every agency: once an administrator turns it on, it automatically flags unusual search patterns for supervisory review and requires a case number and offense type on every search. Flock has said the tool becomes mandatory for all law-enforcement customers on January 1, 2027. Next door, the Brunswick County Sheriff’s Office event log shows the exact moment it was switched on — a line reading “Audit Assistance: Disabled → Enabled,” dated August 18, 2026 (we documented that). Wrightsville Beach’s event log contains no such line. Across every entry the Town produced, there is no record of it ever enabling Audit Assistance.

That gap matters, because Audit Assistance is the safeguard built to catch exactly the kind of misuse a plate-reader program can hide. The Town has said an audit caught a former officer — Tonya Baker, indicted September 8, 2026 for allegedly misusing a reader between February and May 2026 (an accusation, not a conviction) — yet the alleged conduct ran for months, and no charge came until September. Whether an always-on automated audit would have surfaced that pattern sooner is precisely the question the Town’s withheld records would answer. We cannot prove from a produced log that simply omits it whether or when the Town turned the tool on; the organization-settings history and the per-search Organization Audit would. But the contrast with Brunswick — whose log shows the switch being thrown — is hard to ignore. (More on the Town’s cameras and program: our Wrightsville Beach page.)

One hundred days, letter by letter

Every document below is posted in full. Click any letter to read it.

Pink rows are the Town’s shifting positions. Every document is posted in full at the Internet Archive and tracked in our records repository.

The same nine-item request in every row. Four of the six responding agencies produced records within eleven days; the median was about eleven. Even the slowest to finish — Pender County — closed its request in 35 days. Wrightsville Beach has now spent more than one hundred and still withholds the search logs. Agencies differ in size and records, so read the two beach-town rows — same day, same kind of town — as the closest comparison. Full correspondence is in our records repository.

How long does it take to produce one month of Flock audits?

About as long as it took you to read this far. And we don’t have to guess at that, because Flock documents the process itself.

The records at the center of our request — the Organization Audit (the agency’s own searches) and the Network Audit (outside agencies searching the town’s cameras) — are not compiled by hand from scattered paper files. They are live exports from Flock’s own dashboard. In its Guide to Flock Safety Data for Open Records Laws (September 3, 2025), the company tells agencies exactly where the logs are and how to pull them: they live on the “Insights” tab and can be downloaded “in the user interface as spreadsheets in 31-day increments.”

Read that last phrase again, because it settles the question. “31-day increments” means one month per file. To pull a single month, an authorized user opens the Insights tab, sets the date range, and clicks download. What comes back is a dashboard export — not a record assembled by hand from paper files. It is the same kind of one-click export a store manager runs for last month’s sales.

So the arithmetic of our request is not hard. We asked for three months — March, April, and May 2026 — of the Organization Audit and the Network Audit. At one download per 31-day increment, that is three monthly exports per log: a few minutes of pointing and clicking, on files that already exist. Peer agencies have proven it: NC A&T’s Network Audit came back 1,390,776 rows in eleven days, and UNC School of the Arts exported the same request in three. A clerk who wanted to answer could have the export files generated before a cup of coffee got cold — legal review, where any is genuinely needed, is a separate step. Wrightsville Beach instead took 45 days to produce anything — and when it did, on August 14, it produced the contract, the policy, the sharing config, and an account Event Log, but not the two per-search audits this arithmetic is about. Those remain outstanding.

Which raises the obvious question: if generating the files is that quick, why the wall? The same Flock guide answers that too — not with a technical reason, but with a strategy. It coaches agencies to consider narrowing the timeframe, to weigh charging the requester an up-front fee, and to redact fields before release. The stall is not a limit of the software. It is a playbook, written down by the vendor — and delay is the first play.

The next move is just as predictable: redaction takes time. It barely does. North Carolina shields the plate number, and a handful of identifying or investigative fields may be exempt — but redacting a spreadsheet column is not a research project. It is a find-and-replace: highlight the column, type ***, fill down. Seconds per field, on a file that already exists. (In the Network Audit — the log of outside agencies searching the town’s cameras — those fields are commonly produced already showing *** to begin with.) And nothing about redacting three spreadsheets justifies producing zero of them: a partial production now, with a one-line note for anything withheld, is exactly what § 132-6 contemplates — and exactly what Kure Beach and Carolina Beach delivered within days.

The Town’s own summary shows a prohibited use running second

When counsel wrote on October 7 to say the Town need not “create” records, it enclosed one new document anyway: a Flock offense-type usage summary — a ranked count of how the cameras have actually been searched. It is not the per-search audit we asked for (it names no user, no date, no case number), but it is revealing all the same, because it is the Town’s own number on its own letterhead’s attachment.

The top category is “Welfare Check” at 19.70%. The second is “Traffic Infraction” — 144 searches, 18.70% of all use. That is the problem in a single line. North Carolina law restricts automated plate readers to specific law-enforcement purposes and § 20-183.31(b) prohibits using them for routine traffic enforcement — and the Town’s own Policy 5.09 says the same. Yet the Town’s own summary shows traffic enforcement as the second-most-common use of the system. For comparison, the category the program is always sold on — missing and endangered persons — accounts for just 3.20% (25 searches). The document meant to reassure us is the strongest argument yet for seeing the full per-search log.

It also matters who else may have reached into these cameras. A former Wrightsville Beach officer has been criminally charged with misusing a plate reader — an accusation, not a conviction — over conduct alleged between February and May 2026. The audit period we requested, March through May 2026, overlaps that window almost exactly. The per-search Organization Audit and Network Audit are precisely the records that would show such access, legitimate or not. A town that had nothing to explain would produce them. This one has spent a hundred days explaining why it won’t.

A volunteer map already shows the cameras the town won’t count

Here is the part that turns the delay from frustrating to revealing. One item of our request simply asked how many ALPR cameras the town operates, and where. Wrightsville Beach won’t say. A crowdsourced, volunteer-built map already can.

DeFlock’s public ALPR map — built by OpenStreetMap contributors, not by any government — shows four Flock Safety cameras in Wrightsville Beach (and about six across the immediate area, every one of them Flock). Their placement is the point. Wrightsville Beach is a barrier island with one road on and off: the Causeway Drive drawbridge. The map shows readers at that bridge, on North Lumina Avenue, and at the South Lumina / Waynick beach entrance. In a town with a single bridge, readers positioned at the chokepoints — if active and oriented as mapped — are placed to capture a very large share of every vehicle that enters or leaves Wrightsville Beach.

Four Flock Safety readers, all coded “Traffic.” Two guard the Causeway Drive drawbridge — the island’s only road on or off — with the others on North Lumina and at the South Lumina / Waynick beach entrance. Locations from the volunteer DeFlock / OpenStreetMap dataset (August 2026): a crowdsourced floor, not an official count — which is the very record Wrightsville Beach won’t produce. Tap a marker for its coordinates.

A resident with a phone mapped those cameras. The town whose jurisdiction they sit in — and that we asked, in writing, five weeks ago — still won’t confirm the count, or even whether it owns the readers or merely taps into them — which also means no resident can know whether there are more readers the map has not caught, on town poles or private property. When a volunteer can answer a question the government won’t, the silence starts to look like a choice.

So what is Wrightsville Beach hiding?

Maybe nothing. Maybe the contract is ordinary, the sharing is limited, and the audits are clean. If so, the fix is easy and the model is next door: produce the records, the way Carolina Beach and Kure Beach did. Publish the contract. Publish the policy. Generate the audit exports — minutes of clicking — and produce the plainly separable records now, the way the neighbors did, so residents can see who has been searching the cameras at their town’s only bridge, and why.

Until then, the contrast writes itself. Same request, same day: the neighbors answered in days; Wrightsville Beach answered with a law firm, three shifting excuses, and a hundred days of delay. We have set a final date of October 14, 2026 to produce the Organization Audit and the Network Audit in native, machine-readable form (CSV) — or to say plainly that they will not be produced. A preservation hold under N.C. Gen. Stat. § 132-3 is in force, and the Town’s costs and our attorney’s fees are recoverable under § 132-9 if this has to be decided in court. We will report exactly what arrives, and what doesn’t.

Turnaround figures are from DeFlockILM’s own public-records correspondence, tracked in our records repository. Camera counts and locations are from the volunteer DeFlock map (OpenStreetMap data) as of August 2026 and are a floor, not an official count — which is precisely the record we requested. The Flock export description is quoted from Flock Safety’s “Guide to Flock Safety Data for Open Records Laws” (Sept. 3, 2025). The offense-type figures are the Town’s own, enclosed with its counsel’s October 7, 2026 letter. All correspondence quoted here is posted in full at the Internet Archive. Characterizations of the Town’s position are our opinion based on that correspondence; the indictment of a former officer is an accusation, not a conviction. This page is general information and opinion, not legal advice. First published August 5, 2026; updated October 8, 2026.