Talking Points Against Flock AI Cameras
Walk into any conversation, council meeting, or commissioner’s inbox with the evidence. Different speakers should carry different points — pick the one that’s yours, say it in your own words, and sit down. Every point below is documented and linked to the record. (Every figure, sourced, in one place.)
Pick one lane and lead with it. If you want the three that move a vote:
- The ask. Nobody voted for this, and the Board holds the only lever — stop funding it, the way Macon County did 5–0.
- The secret. The County won’t say who ran 2.98 million searches of our plates. Five other NC agencies produced that field in about 18 days; New Hanover is at 97 and counting.
- The proof gap. The Sheriff says it solves crime, but the County has produced no data showing it does — and North Carolina’s own protections don’t follow your plate across a state line.
The full menu is below. Each point links to the record behind it.
1 Nobody voted for this — and the Board holds the only lever that ends it
No referendum. No public hearing. No commissioner ran on it. And it ends the same way Macon County’s did: the Board stops paying for it.
- The Sheriff’s Office holds Flock contract No. 25-0364, roughly $219,000. (The contract, in full.)
- The Board of Commissioners appropriates that money every year. Macon County’s board voted 5–0 on July 15, 2026 to stop appropriating theirs — it did not cancel a contract; it opted out at renewal and voted the cameras down. (WLOS)
- New Hanover’s contract also lets the County cancel at any time, without cause. (The clause, quoted.)
A commission holds the money. That is the whole mechanism — and it is the ask. (The contract and the cost · where each commissioner stands.)
2 The County will not say who searched — five other NC agencies did, in days
This is the question everything else runs through, and it is unanswered.
- The County produced the audit with the name of every searching agency blacked out — 2,980,082 rows of ***.
- Five other North Carolina agencies on the same vendor produced that exact field — naming the searching agencies while still protecting individual officers: Carolina Beach, NC A&T, UNC Pembroke, Kure Beach, and UNC Charlotte police, on average in 18.2 days; NC A&T in eleven. Buncombe County released it too.
- New Hanover has withheld it for 97 days and counting. We filed a request for pre-litigation mediation on August 3, 2026. (The full timeline, the comparison, and the filing.)
- And it is not only the County. We sent the same records request to three beach towns on the same day — Carolina Beach and Kure Beach produced within days, but Wrightsville Beach held out 45 days, referring the matter to outside counsel, before producing on Aug 14 — and it still withholds the audit logs naming who searches its cameras. The pattern repeats wherever the public tries to see who is searching. (Wrightsville Beach’s 45 days, documented.)
Ask your commissioner one question: who searched our cameras, and why won’t we say? (The correspondence, in full.)
3 To one NC agency, every Flock record is a ‘criminal investigation’ — which means everyone is a suspect
Asked to produce its Flock logs, a North Carolina police department gave an answer that says the quiet part out loud:
- Durham police have taken the position that their Flock usage logs “constitute a record of a criminal investigation … and are not public records.” (INDY Week)
- Read it literally and it is an admission: if every search of the plate network is part of “a criminal investigation,” then every driver it logs is being treated as the subject of one — which is precisely the objection to mass ALPR. Read it cynically, it is a label pinned on after the fact to keep the public from seeing how the system is used. Either way, it damns the tool.
- It is also a stonewalling roadmap: an agency that calls all ALPR data “criminal investigation records” can try to deny every public-records request for it under § 132-1.4 — the same confidentiality New Hanover is leaning on to hide who searched.
A machine that treats everyone as a suspect — described, in writing, by the people who run it. (What to do when an agency says no.)
4 It has not been shown to work — and the County has produced no proof that it does
The Sheriff’s whole case is that it solves crime. The evidence for that is thin, and the County has produced none of its own.
- In the SBI’s own state pilot, Raleigh’s three state-road cameras recorded 14.8 million drivers and returned 956 “hits” — the report never says whether a single one produced an arrest. (Separately, the Daily Tar Heel reports the same three cameras generated 227 stolen-vehicle alerts — but an alert is not a recovery or an arrest, and neither the SBI report nor the reporting says how many led to one.) (The SBI report to the General Assembly.)
- We asked New Hanover for any study, data, dashboard, or “impact” material showing the program works. That request remains unanswered. (The records timeline.)
- Flock’s claim that it helps solve ten percent of U.S. crime rests on a company-run study whose own named academic reviewer has publicly questioned it. (Flock’s arguments, answered.)
“It solves crime” is an assertion, not a finding. Ask to see the number. (The efficacy claims, examined.)
5 North Carolina’s protections don’t travel with your plate
North Carolina writes real limits into its ALPR law. None of them bind the agency in another state searching New Hanover’s cameras.
- North Carolina bars the sale of captured plate data, caps retention at 90 days, and bars using it for traffic enforcement (§ 20-183.31–.32).
- The busiest searchers of the shared network sit in Texas, Georgia, and Florida — states with weaker rules or none. Texas has no plate-reader law at all, and nothing in Texas law stops a sale. (Where New Hanover’s reads reach.)
- The protections North Carolina writes for itself do not follow your plate across a state line. The County exported its residents’ movement data into legal regimes its own legislature rejected. (The nationwide picture.)
A limit that stops at the state line is not much of a limit. (Who is searching North Carolina’s cameras.)
6 The Sheriff enrolled New Hanover in a national surveillance network — on the taxpayers’ dime
This is not a local camera system. It is a node in a nationwide network, paid for by New Hanover residents.
- Aggregated records from five North Carolina agencies on the network show 3.9 million searches in three months — the large majority run by agencies outside North Carolina. (The aggregate analysis.)
- Agencies in other states can reach New Hanover’s cameras, and no one in North Carolina reviews their searches or verifies they are legitimate. (Who can search our cameras.)
- Local taxpayers fund the hardware; out-of-state and federal agencies get the reach at no charge. (The network, mapped.)
We paid to be watched — and to watch strangers we will never meet. (Who is searching North Carolina’s cameras.)
7 It searches for people, not just plates
The pitch is license plates. That is not the whole product. New Hanover County’s own audit log records the other half, and it is the fact most commissioners have never been told:
- 470 “FreeForm” searches ran against this county’s camera network between April 2025 and May 2026. FreeForm lets an officer skip the plate and type a plain-English description instead.
- 456 described vehicles. Fourteen described human beings — clothing, build, what someone was standing near. (Our row-by-row breakdown.)
- Three of the fourteen typed a race, and in each case the identical search was re-run seconds later with the race removed. We think that is Flock’s guardrail engaging. (The three pairs, with timestamps.)
- We cannot tell you who ran any of them. The County redacted the searching agency on all 2,980,082 rows.
Somebody typed “2 young teenage black males” into a people-search against cameras this county pays for. Flock’s filter blocked that one; fifteen seconds later the same search with the race word removed was allowed. (The searches, in full · Sheriff McMahon called it “just a license plate reader” — watch this.)
8 The system flags cars before any officer suspects anyone
Hotlists and pattern tools run without a human typing anything. They are in this county’s records too:
- 4,955 “convoy” searches touched this county’s network — a paid feature that asks which vehicles habitually travel with a given car. (How convoy search works.)
- Down the road in Kure Beach, whose production names the searching agency, 19 convoy searches were run by the N.C. State Bureau of Investigation in three months. (Kure Beach’s audit.)
- Flock’s Multi-State Insights alerts police when a vehicle is seen in more than one state. Crossing a state line is now a signal. (Our explainer.)
- A custom hotlist lets an officer add a plate that then trips an alert on every pass — Kure Beach’s event log shows 21 such entries in three months, in a town with four cameras.
The Fourth Amendment puts suspicion before surveillance. This reverses the order. (How hotlists and pattern flags work.)
9 It maps the pattern of your life
Plate data is not a list of cars. Assembled over months it is a record of where you go and who you are:
- The church you attend, the clinic you visit, the range you shoot at, the union hall, the recovery meeting, the lawyer’s office, the protest.
- WRAL obtained the internal ALPR policies of all 32 agencies in the SBI pilot. Only Asheville and Raleigh police restrict using readers to surveil political rallies, religious gatherings or protests. The other thirty have no such language. (The context.)
- Gun owners: every trip to the range or the gun shop is logged, without a warrant and without a crime. (What it logs about gun owners.)
Where you have been, and who you were with, becomes searchable by strangers. (Watched without a warrant.)
10 The contract gives the vendor access the Sheriff’s own policy forbids
The Sheriff’s policy says only his employees may touch the data. The contract he signed says otherwise.
- SOP 222 restricts access to Sheriff’s Office employees and logs every one of their searches by login and password.
- The Flock contract gives the vendor — a private company — standing rights to access, use, and disclose the footage to third parties on its own judgment, and to train AI on the data. Nothing requires Flock to log its access the way the County logs its own staff. (The clauses, side by side.)
In our reading that sits in tension with state law; regardless, it contradicts the Sheriff’s own policy. A commissioner can ask whether any log of vendor access exists. (The policy vs. the contract.)
11 Nobody audits any of this from outside
Every accountability claim runs through the audit log. Here is who checks it:
- § 20-183.31(a)(7) requires only an annual audit reported “to the head of the agency” — not to a board, not to a court, not to the public. Once a year, to the boss.
- There is no external audit, and no audit of the system’s security or sharing settings at all. Asked in June 2026 whether Flock had ever been independently audited, the Southern Pines police chief said he was unaware of any. (Sandhills Sentinel)
- The only penalty in the statute for misuse is a Class 1 misdemeanour, and it reaches plate data only.
A system that watches everyone and is reviewed by no one is not oversight. It is paperwork. (Who watches the watchers.)
12 The people who hold the keys keep misusing them
This is not a hypothetical risk. It is a documented pattern, and it is accelerating — including in North Carolina:
- On July 29, 2026 the SBI charged a Charlotte-Mecklenburg police officer — arrested the next day, July 30 — with accessing computers (a misdemeanor under N.C.G.S. § 14-454(b)), for allegedly running an unauthorized Flock search of a North Carolina plate. The department owns no Flock cameras of its own; the officer reached into other agencies’ cameras through a data-sharing agreement. The charge is an accusation, not a conviction. (The North Carolina charge.)
- Then, a week later — August 5, 2026 — a Mooresville, NC officer was dismissed and charged — with a misdemeanor, accessing computers — for running the Flock database 31 times to stalk a private citizen. Her department’s chief said ten more officers there are under investigation for personal-use access, three of the cases “problematic,” and that another arrest could come “within days.” That is two North Carolina officers criminally charged in a week — both with misdemeanor computer-access crimes — plus ten more under review in one department. (Caught by the audit — the NC misuse cases.)
- Five former Albany, Georgia officers were arrested on 21 counts for misusing plate-reader data for personal reasons. (GBI)
- In Georgia — Flock’s home state — more than a dozen officers across at least nine agencies were fired or charged in 2025–2026. A Milwaukee officer ran 179 Flock searches on an ex-partner. (The full tally, case by case.)
The question is never whether the tool is misused. It is how long it takes anyone to notice. (The stalking cases.)
13 Federal agencies get in — through the side door
Flock committed in August 2025 that “federal users will not be added to Statewide or Nationwide lookup.” That covered the lookup pools. It did not cover direct shares:
- Kure Beach shares its network directly with the FBI and the U.S. Postal Inspection Service — the share runs one way. Those two agencies ran 75,550 searches of that town’s cameras over five months, well after Flock’s commitment. (The numbers, broken out.)
- Nationally, Flock’s CEO told the public the company had “no federal contracts” while four federal agencies — including the investigative arm of ICE — held pilot access, and an Illinois audit found it broke state law. (What the records show.)
The safeguard is a setting. A four-camera beach town can change a setting. (Who can search our cameras.)
14 It is wrong often enough to put innocent people in handcuffs
Errors here are structural, not occasional — and North Carolinians have already paid for them:
- The Institute for Justice has catalogued 27 wrongful-stop cases since 2018 — innocent motorists pulled over, held at gunpoint, or jailed because the camera or database was wrong. In nearly two-thirds, officers did not realise the error until after they had drawn their guns. (The cases.)
- Two North Carolina women were wrongly arrested after readers flagged the wrong car. (The North Carolina cases.)
- Settlements have run from $45,000 to $1.9 million (Aurora, CO — a mother and four children face-down on the pavement). In Oak Park, Illinois, oversight found a third or more of Flock-prompted stops ended with the driver released over data problems. (Sources and detail.)
A machine turns one truncated database entry into four squad cars around an innocent person. (The wrongful-arrest record.)
15 Other officials — including police chiefs — no longer trust the company
The people saying Flock cannot be trusted are not activists:
- A California police chief who had vouched for the cameras shut off all thirty and said he “personally no longer ha[s] confidence in this particular vendor.” (His statement, and the staff report behind it.)
- A North Carolina judge found Flock operated in the state without a license after being ordered to stop; an Illinois state audit found it broke state law; a U.S. senator concluded abuses are “inevitable.” (The court order, the audit, the letter.)
- Denver councilmembers, in writing: the CEO “has made multiple false statements … which raises significant questions about Flock’s trustworthiness as a vendor.” (The cancellations, in their own words.)
When customers walk away citing trust — police chiefs among them — that is evidence. (What Flock told its customers, and what the records show.)
16 The logs show traffic enforcement — which North Carolina law forbids
State law bars ALPR for traffic enforcement (§ 20-183.31(b)). The records next door show it happening anyway:
- Kure Beach’s own audit — which names the reason for each search — shows 275 searches, about a third, coded “Traffic Infraction,” a use North Carolina law prohibits. (The reason field, analysed.)
- Carolina Beach’s records show it too. The Town’s own officers coded 41 searches “Traffic Infraction” (and 15 more “Traffic Analysis”) — against the Town’s own Policy 427.4(g), which repeats the statutory ban — and, once produced unredacted, the network audit showed 258,521 “Traffic Infraction” searches run against Carolina Beach’s cameras by outside agencies. (The Carolina Beach production.)
- New Hanover’s own network audit shows 17,695 searches coded “Traffic Infraction” among the 2.98 million run against its cameras. The County blacked out who ran them — but not the stated reason. (What the County withholds.)
The statute is categorical, and New Hanover’s own record shows the barred use — 17,695 times. (Kure Beach’s records.)
17 “It was mislabeled” is not the escape it sounds like
Shown its own Flock audit coding searches as “Traffic Infraction” — a use § 20-183.31(b) forbids — the New Hanover County Sheriff’s Office told WECT (Aug 4, 2026) the label was a mistake. We take that seriously. Here is the fair answer:
- The label is the agency’s own. “Traffic Infraction” is a reason an officer picked from Flock’s own dropdown, on the agency’s own system, recorded in the agency’s own audit. We are quoting the record, not inferring a purpose.
- The statute has no threshold. § 20-183.31(b) does not bar “mostly” traffic enforcement, or traffic use above some percentage — it is categorical. A miscoding defense has to establish the data was not used for traffic enforcement at all; the agency’s own record says otherwise, repeatedly.
- If the labels can’t be trusted, that is the bigger problem. Agencies and Flock point to the audit log as the safeguard that catches misuse. An agency that says its own audit labels are unreliable is conceding its one accountability mechanism doesn’t reliably record why a search was run. An audit you have to disclaim is not oversight.
The fix belongs to the agency: correct the labels, produce what each “Traffic Infraction” search was actually for, and confirm in writing that no ALPR data is used for traffic enforcement. Do that and we will publish it, prominently. Until then, the agency’s own record is the record. (The reason field, analysed.)
18 Why is a Sheriff’s camera pointed at a children’s soccer field?
New Hanover’s Flock contract also bought hidden “Condor” pan-tilt-zoom video cameras — not plate readers. The County won’t say where they are. We found one:
- A Condor live-video camera sits at Veterans Park, pointed at the youth soccer fields — we published the photo. (See it, and how to spot the others.)
- These cameras pan, tilt, zoom, and stream live. Eight are in the contract; the County redacts their locations. (The Condor order, quoted.)
A commissioner can ask, from the dais, where the other seven are. (The Veterans Park camera.)
19 A second North Carolina misuse record: UNC Pembroke
This is not only a Georgia problem. A North Carolina university’s own audit shows the same pattern:
- UNC Pembroke’s Flock network audit — roughly 1.5 million searches in three months — shows the same overwhelming out-of-state reach into a small campus system. (The UNCP audit.)
The pattern crosses state lines, and it is already inside North Carolina. (UNC Pembroke’s records.)
20 The law is moving — and the record here is bigger than the cases that permitted it
Courts are working this out now, and we will state it carefully:
- In Carpenter (2018) and again in Chatrie (2026), the Supreme Court has narrowed warrantless location tracking.
- In United States v. Porter (5th Cir., 2026), a court held plate-reader use was not a search — distinguishing Carpenter partly because that town had “no more than ten LPR cameras.”
- New Hanover County has roughly 28, wired into a network searched 2.98 million times in sixteen months. Whether ten cameras and a nationwide network are the same constitutional question has not been decided on a record like ours.
We are not predicting an outcome. We are saying the ground is moving, and the County is standing on it. (ALPR in the courts.)
21 Flock is not the only thing watching Wilmington
Cancelling one contract does not end the inventory. Residents should know the whole of it:
- Wilmington PD runs a Fusus “STING Center” joining public and private cameras into one live feed; the city also runs ShotSpotter and drones. Separately, the Sheriff’s Office — not the city — bought ShadowDragon social-media surveillance.
- Onslow County runs Plate Ranger, a different vendor entirely — the same questions apply to all of them. (The inventory.)
The point is not one company. It is that nobody voted for any of it. (The full surveillance inventory.)
22 North Carolina towns have already said no — by vote, in public
This is not a fringe position. It is what neighbouring boards concluded once they read their own contracts:
- Macon County — commissioners voted 5–0 to defund and remove the cameras. Hillsborough — unanimous cancellation over Flock’s data-sharing terms. Pittsboro — voted 4–1 to terminate. Chatham County pulled its own cameras.
- Apex, Garner, Asheville, Southern Pines and Madison County are contested now. (The statewide tracker.)
Every one of those started where we are: a signed contract and a worried public. (The statewide tracker.)
23 A free nation does not track innocent citizens
Strip away the statutes and the audits, and one question remains:
- This system treats every driver as a suspect to be logged, indexed, and made searchable — without a warrant and without a crime. (Watched without a warrant.)
- The evidence it helps is anecdotal. The evidence that it watches everyone is in the County’s own records. This is the surveillance we once said only happened somewhere else. (America at 250.)
A free nation does not track innocent citizens. That is the whole of it. (The Fourth Amendment case · the conservative case.)
Our goal is removal — because regulation can’t fix this.
We’re honest about the ask: cancel the contract and take the cameras down. People ask, “why not just regulate it?” We used to list the safeguards a responsible program would have. We don’t anymore — because every one of them fails in practice. Here is that same list, and why each so-called guardrail doesn’t hold:
- 1A “transparency portal” shows you nothing that matters. New Hanover’s went live in August — and still hides the camera locations and who has searched the system. A portal is a dashboard, not oversight.
- 2A warrant requirement is unenforceable here. No judge ever sees these searches; the only review North Carolina requires is the agency auditing itself, once a year. A rule no one checks is not a rule.
- 3“No out-of-state sharing” is a setting the vendor can flip back. Nationwide access is a toggle — about 3,904 agencies, running roughly 90% of the searches from outside North Carolina, already reach these cameras. A guardrail you can override is not a guardrail.
- 4Retention limits are just another overridable default. The Sheriff set 30 days, then said within days he was “reevaluating” it; the vendor moved its own default to 7. Numbers that change on a whim protect no one.
- 5Audits don’t stop misuse — they only document it after the fact. A Mooresville officer ran 31 searches to stalk a private citizen; the audit logged every one and prevented none. And the reason field an audit reviews is self-entered — New Hanover already said its own codes were “mislabeled.”
We say this plainly: you cannot regulate a warrantless dragnet into something that honors a warrant. Every “safeguard” above is a setting, a promise, or a paperwork step — overridable, unenforceable, or after-the-fact. The one remedy that actually works is the one the Board can order today: cancel the contract and take the cameras down. (The full case: why regulation and “guardrails” don’t work.)
And we have already asked. The three challengers on the ballot all oppose Flock; the incumbents running for re-election have taken no official stand; most of our letters went unanswered; and the contract still is not on any agenda. When the people who could regulate this won’t even answer, removal is the only lever left.
Turn facts into pressure
Sign & share
Sign the petition opposing Flock and similar ALPR networks, then send it to a neighbor.
Talk to neighbors
Raise the privacy implications with neighbors and community organizations. Grab a flyer.
Demand accountability
Attend city council and county commission meetings and email your leaders to demand removal.
Privacy is not a luxury — it’s a right.
“Ultimately, saying that you don’t care about privacy because you have nothing to hide is no different from saying you don’t care about freedom of speech because you have nothing to say.”— Edward Snowden
You’re not a suspect. So stop being tracked like one.
It takes one minute. Add your name — then tell the officials who represent you to take the cameras down.