Where You Drove. What You Spent. Who Gets to Ask.

Where the line sits — the whole argument in one table:

The left column is documented and cited below. The middle column is technically possible, not shown to be happening. The right column is what this article does not claim exists. Keep those three apart and the rest follows.

Part One — The Local Record

The number is not a projection. It is a receipt.

Between January 2025 and June 2026 — sixteen months — the Flock network was searched roughly 2.98 million times against data collected in New Hanover County, per WECT’s analysis of the county’s records. The county holds about 230,000 residents. A separate WECT tally found the Sheriff’s Office ran only about 6,000 searches of its own between June 1 and early August 2026. In the audit the county produced, the searching agency and officer are redacted — so the records do not identify who ran the rest.

SEARCH LOG — NEW HANOVER COUNTY / FLOCK NETWORK
DATE 2025-12-11 · AGENCY [REDACTED] · OFFICER [REDACTED]
REASON: “2 young teenage Black males”
RESULT: BLOCKED
FOLLOW-UP: descriptor removed → search ALLOWED
Reported by WECT Investigates, August 18, 2026. A second instance on February 4, 2026 followed the same pattern. The guardrail filtered the wording. It did not stop the search.

The contract is #25-0364. Roughly $219,000. Signed February 2025, running to 2027. It covers 36 cameras, of which 19 read license plates; the balance includes solar Falcon units and long-range cameras. The Wilmington Police Department is not a party to it. This is the New Hanover County Sheriff’s Office.

On August 18, roughly 41 residents spoke during public comment at the county commissioners’ meeting. Not one spoke in favor. Commissioners took no vote and committed to reviewing the contract.

“I don’t believe that any of us were aware of what the contract had to say.”
— Commissioner Rob Zapple, New Hanover County Board of Commissioners, August 18, 2026

Read that again as a procurement fact rather than a political one. The contract was public the entire time. The audit logs were obtainable the entire time. What changed in August was not the availability of the record. What changed is that someone requested it, read it, and brought it into the room.

That is the whole method, and it is available to any resident of this county.

Part Two — What the Hardware Actually Is

The camera on your street is a terminal, not a system.

A Flock camera is solar-powered, pole-mounted, and connected by commercial cellular backhaul rather than municipal wiring. A hardware teardown published by the Center for Human Rights and Privacy documented an LTE modem inside the unit. That detail carries more weight than it appears to: the camera does not need the county’s network, the county’s fiber, or the county’s permission to reach the cloud. It needs a pole and a signal.

Modern readers do more than transcribe plates. They extract a vehicle description — make, model, color, roof rack, damage, bumper stickers — which is why a covered or missing plate degrades identification without defeating it.

Every capture travels to a privately owned national database. As of August 2026, reporting places the Flock network at more than 120,000 cameras across 49 states, serving upward of 5,000 law enforcement agencies. The company states its cameras produce roughly 20 billion plate scans a month. That figure is self-reported and not independently audited; we cite it as the company’s claim, because it is.

The number that matters locally is smaller and harder to argue with. A single search run against New Hanover County data can reach up to 7,000 networks nationwide. The county bought 36 cameras. What it also acquired was access to a national network.

And the 36-camera Flock contract is not even the county’s whole camera footprint. The Sheriff’s Office separately runs “Connect New Hanover Sheriff,” an Axon Fusus program whose own public site reports 2,147 registered cameras and 716 already integrated for direct live-feed access — private and business cameras, wired into a county platform with AI that can “automatically recognize weapons, vehicles of interest”. The Flock contract is the camera program with a public debate. By camera count, it is also the smallest of the county’s camera programs — though the two systems differ in ownership, access, and function, so counts alone don’t settle which reaches furthest. The plate readers log where you drove; the Fusus network watches where you are — and neither went before the public in a separate, recorded vote.

Part Three — The Documented Record

The case against dragnet ALPR does not rest on prediction. It rests on audit logs.

Every entry below is documented in reporting, sworn filings, or a state audit. Nothing here requires anyone to speculate about intent.

2025-05-09 · Johnson County, TX — A sheriff’s deputy searched 83,345 cameras for a woman who had an abortion.
Two national queries: the first across 1,295 networks and 17,684 cameras, the second across 6,809 networks and 83,345 cameras. The audit-log reason field read “had an abortion, search for female.” The sheriff characterized it as a welfare check by a worried family. Court records and sworn affidavits later obtained by EFF show deputies simultaneously opened a death investigation into a non-viable fetus and consulted prosecutors about charges. She was not charged. Both things are in the record.
404 Media, May 29, 2025 · EFF, October 7, 2025

2025-05-27 · Danville, IL records — Local police ran more than 4,000 lookups for federal immigration purposes.
A single public records request to one Illinois police department exposed nationwide search logs. Reason fields included “immigration,” “ICE,” “ICE+ERO,” and “ICE WARRANT.” Every lookup explicitly citing immigration post-dated January 2025. The structural finding is the one to hold onto: Flock has no contract with ICE. The agency obtains the data by asking local police to run the search. That mechanism routes around state law without violating any contract.
404 Media, May 27, 2025

2025-08-25 · Illinois Secretary of State — A state audit found the vendor in violation of state law.
Illinois amended its expressway camera act in 2023 to bar ALPR data sharing for immigration or abortion enforcement. The audit found Flock had given Customs and Border Protection access to Illinois data through a pilot program, and that out-of-state agencies accessed Illinois data without the written declarations the statute requires. Forty-six out-of-state agencies were cut off. Flock’s CEO conceded the company “clearly communicated poorly” and had not created distinct permissions and protocols. Note what the statute did not contain: a penalty provision.
Illinois SOS, August 25, 2025 · NPR Illinois, August 27, 2025

2025-12-22 · Nationwide — At least 60 cameras were left livestreaming on the open internet with no password.
Reporters could watch live feeds, download 30 days of archived video, alter settings, and read system logs — without credentials. They tracked themselves across multiple cameras. The units involved were Condor models, a pan-tilt-zoom camera built to record and track people rather than vehicles, with automatic zoom onto faces.
404 Media, December 22, 2025

2026-08-02 · National — At least 50 officers were accused or charged with misusing plate-reader networks.
Flock’s system was involved in 46 of those cases. Named instances include a Georgia police chief who ran roughly 600 queries over more than a year tracking an ex-partner, a Kansas chief fired for tracking an ex, and a Milwaukee officer who tracked a partner roughly 180 times in two months. The structural finding again matters more than any single case: audit logs are reviewed after a complaint triggers an investigation. They are not reviewed proactively. The departments were not looking. A newspaper looked instead. (The North Carolina misuse record.)
Washington Post investigation, August 2, 2026 · follow-up reporting, August 13 and 19, 2026

Flock announced remediation on August 13, 2026 — one day after the first Post follow-up. The changes are real and worth acknowledging: recommended retention cut from 30 days to seven, mandatory case codes by year-end, offense filtering to limit which outside agencies can search, automatic suspension of users flagged for abnormal activity. Sheriff Ed McMahon announced his own changes on August 10: retention cut from 90 days to 30, audits moved from annual to monthly, and a public transparency portal.

Both responses share a gap. Neither addresses national lookup. The 2.98 million searches against this county’s data were overwhelmingly not run by this county.

Part Four — North Carolina Law

The statute is real, narrow, and quieter than most residents assume.

N.C.G.S. Chapter 20, Article 3D — §§ 20-183.30 to 20-183.33

The retention debate in this county — 90 days, then 30, with Flock now recommending seven — is a debate about how long the record persists inside Flock’s own system. It is not a debate about who may search it or on what basis. And a shorter retention window guarantees far less than it sounds: nothing in it prevents the data from being exported, copied, or shared elsewhere before it is deleted, and nothing requires an independent, external audit to confirm the deletion ever happens. A record you cannot see, cannot follow once it leaves the box, and cannot verify was destroyed is not made safe by a countdown timer. Those are different questions, and only the least important one — the timer — has been touched.

The state moved in July, and it moved toward more cameras

S.L. 2023-151 created a pilot program permitting the State Bureau of Investigation to place readers on state-controlled rights-of-way — roughly 80 percent of North Carolina roadway mileage — coordinated with the Department of Transportation. House Bill 206 would have made the pilot permanent and stalled in Senate Rules. The provision was enacted anyway, through the state budget signed July 7, 2026. Each installation requires NCDOT approval before placement. Reporting placed the program at roughly 140 camera locations as of March 2026, with about 75 more pending approval or installation.

What Austin learned. Austin’s city contract with Flock ended June 30, 2025. In February 2026, the Texas Department of Public Safety installed state-operated Flock readers on state rights-of-way inside Austin — including outside City Hall — without initially disclosing the locations. A local cancellation removes a local contract. It does not remove a state camera from a state road. North Carolina now has the same structure Texas used. Any strategy that ends at the county line is incomplete by design.

Part Five — The Legal Ground

The Fourth Amendment moved in June. It has not arrived here yet.

Three decisions define the terrain.

In United States v. Jones (2012), the Supreme Court held that installing a GPS tracker on a vehicle and monitoring it for 28 days was a search. Justice Sotomayor, concurring alone, wrote that GPS monitoring “generates a precise, comprehensive record of a person’s public movements that reflects a wealth of detail about her familial, political, professional, religious, and sexual associations.” Her concurrence is not binding law. It is the clearest statement of what aggregated location data reveals.

In Carpenter v. United States (2018), the Court held that acquiring seven days or more of historical cell-site location data is a search requiring a warrant, because a person retains a legitimate expectation of privacy in the record of his physical movements even when a third party holds it. Chief Justice Roberts: “Only the few without cell phones could escape this tireless and absolute surveillance.”

In Chatrie v. United States (No. 25-112), decided June 29, 2026, the Court held 6–3 that the government conducted a Fourth Amendment search when it obtained location data from Google, and rejected the argument that short-duration location data escapes constitutional protection. It did not finally decide whether the warrant itself was reasonable — it sent the case back to the lower court for that. The majority described government access to such databases as “a virtual panopticon with which to scrutinize its citizens’ activities.”

The pending test is Schmidt v. City of Norfolk. On January 27, 2026, a federal district judge granted summary judgment for the city, reasoning that ALPR surveillance “could become too intrusive” but had not yet, because the cameras capture less than the whole of a person’s movements. The Institute for Justice appealed, and the case was docketed at the Fourth Circuit on February 26, 2026 as No. 26-1227. Briefing completed in June. Cato and the New Civil Liberties Alliance filed in support of the plaintiffs.

Note the sequence. The district court ruled in January. Chatrie came down on June 29, while the appeal sat undecided. The Fourth Circuit is positioned to become the first appellate court to weigh a plate-reader dragnet against the Supreme Court’s most recent location-privacy holding. North Carolina sits in the Fourth Circuit, so its ruling would be controlling precedent for the federal courts here on the Fourth Amendment questions it resolves.

Michael Soyfer of the Institute for Justice identified the doctrinal hook that matters: the Chatrie Court “drilled down on what was in the database and not just what police happened to access at a point in time.” That shifts the inquiry away from the individual query and toward the existence of the dragnet itself — which is precisely the question a county commission gets to decide before a court ever reaches it.

Part Six — The Convergence: Physical Tracking and Financial Control

A record of where you went is one half of a permission system.

Here is where this piece stops being only about cameras — without leaving the ground it stands on.

A camera on a roadside pole is not just a camera. It is one piece of a larger system that records movement, stores it, and makes it searchable — and that system has a second half most camera arguments never mention. Put plainly: a camera records where you go; a digital payment system can decide whether your transaction clears. Neither is extraordinary by itself. The debate begins when the two can be connected. The two halves are usually argued in separate rooms, by separate coalitions, on separate schedules. One records where you are and what you do; the other can govern whether a transaction is permitted to complete. My concern — and I’ll state it as a concern, not a finding — is that if those two independently operated systems were ever connected, they could form a single check that participation in ordinary life has to pass. Whether they are joined is the open question this section examines, not a fact I am asserting.

Consider the mechanics. Fixed ALPR networks log ground movement. Automated aerial systems — drone-as-first-responder programs — expand that visibility from fixed points into three dimensions, untethering surveillance from the pole. And digitized financial rails govern whether a purchase — fuel at the pump, a toll, a plane ticket — is allowed to process. This is not distant theory here. Wilmington’s own police department already runs a drone program and a Fusus real-time crime center, and the Sheriff’s Office runs its own Fusus network — 2,147 registered cameras, 716 already streaming live, with AI object detection. The ground layer in this county already has an aerial one bolted on, and an AI watching both.

Be precise about what is and isn’t true today. The systems do not appear to be integrated. But increasingly capable pieces exist on both sides: systems that collect movement data, and systems capable of blocking or freezing transactions. The physical sensors are being mounted on local poles by county contract; and federal law now requires covered payment-stablecoin issuers to build the capability to block, freeze, and reject transactions and to comply with lawful freeze-and-seizure orders. Integration — not invention — is the unresolved step. That is a narrower claim than a control grid already running, and a more unsettling one, because it means the missing piece is a connection, not a breakthrough.

The financial layer is the one most camera arguments never mention, and it is further along than most people realize. The officials describing it have not been coy.

“We don’t know, for example, who’s using a hundred-dollar bill today… A key difference with the CBDC is that the central bank will have absolute control on the rules and regulations that will determine the use of that expression of central bank liability, and also we will have the technology to enforce that.”
— Agustín Carstens, General Manager, Bank for International Settlements, IMF seminar, October 19, 2020

Two years later, at an IMF seminar on central bank digital currencies and financial inclusion, Deputy Managing Director Bo Li — previously a senior official at the People’s Bank of China — described the mechanism directly: programmability lets a currency “allow targeted policy functions,” and “by programming, CBDC money can be precisely targeted for what kind of people can own, and what kind of use this money can be utilized.”

The gap in the law is not where most people look

In Carpenter, the Court declined to extend the third-party doctrine to cell-site location data. It did not overrule United States v. Miller (1976), which held that bank records carry no reasonable expectation of privacy because you voluntarily conveyed them to the bank. Miller remains good law.

Set the two lines of authority side by side and the asymmetry is stark. Your physical movements gained Fourth Amendment search protection in 2018 (Carpenter), and the Court reaffirmed that protection for database location records in 2026 (Chatrie). Your transactions never have. The half of the system that determines whether you can buy fuel, food, or a plane ticket sits on the weaker side of the doctrine, not the stronger one.

That is the argument privacy advocates on both sides of the aisle keep arriving at from opposite directions. The ACLU and the Electronic Frontier Foundation litigate against dragnet location tracking on behalf of protesters, organizers, immigrants, and people seeking reproductive care. Constitutional conservatives and sound-money advocates litigate and legislate against programmable currency on behalf of small business owners, gun buyers, and churches. Different clients. Same infrastructure.

What Congress actually did, and what it did not

Precision here separates a serious argument from a shareable one.

Read the table honestly and it cuts both ways. Banks have always been able to decline, freeze, or report a transaction — under sanctions law, anti-money-laundering rules, court orders, and fraud controls. The question worth raising is not whether a payment can ever be stopped; it is who can impose the restriction, under what authority, with what due process, and using what data. And on the sharpest version — programmable money running against a camera network — no one in the United States is doing that today. The claim that they are is false, and making it costs credibility this campaign cannot spare.

What the table does show is that the freeze on a retail CBDC expires in 2030, while the statutory requirement that stablecoin issuers build freeze-and-reject capability is permanent and already in force. The capability is being constructed on the private side of the ledger while the public-side prohibition runs on a timer. Those are the facts. The inference is left to the reader, which is where it belongs.

The frame, and its limits

The most widely circulated version of this convergence argument belongs to Catherine Austin Fitts, who served as Assistant Secretary of Housing at HUD and Federal Housing Commissioner in 1989–1990 under President George H. W. Bush. She argues that modern societies are building what she calls a “digital control grid,” in which surveillance and financial infrastructure increasingly work together — built not by decree but by ordinary participation: the engineers who write the code, the technicians who mount the poles, the retirement funds that capitalize the vendors, the consumers who trade privacy for convenience. I treat her as a hypothesis-generator, not an authority. I remain skeptical of parts of her account. But after reviewing the court rulings, audits, contracts, and reporting above, many of the underlying technologies she points to are plainly real. The useful question is not whether every one of her predictions is correct. It is whether residents should understand the systems already being built around them.

That structural observation stands on its own and is worth taking seriously. Fitts has also promoted claims about COVID-19 vaccines, the 2020 election, and covert federal spending that fact-checkers and mainstream outlets classify as conspiracy theories. Cite the framework; do not lean the evidence on her. Every load-bearing fact in this article comes from a court record, a state audit, a statute, or a named news investigation — which is the only standard that survives a county commission meeting.

Keep the asymmetry of the clock in view. Federal monetary policy is debated in Washington over years, and no one is running programmable money against a camera network today. But the physical input side of this grid — the cameras in New Hanover County, and the drones already overhead — is not a future debate. It is up for a vote this month. That is the half you can still see, still vote on, and still remove, which is where the rest of this piece returns.

Part Seven — The Signatures

Nothing gets installed on a public road without a name on a form.

A surveillance network is not an abstraction. It is a pole, an encroachment agreement, a licensed installer, a bond, a data plan, a signature page, and a legal opinion advising that all of it is fine. Every one of those artifacts is a public record in North Carolina under Chapter 132.

The state’s July 2026 expansion makes this local. Cameras on NCDOT-owned or maintained rights-of-way require NCDOT approval before placement. That approval generates a paper trail naming the applicant, the installer, the engineer, and the bonding party.

We are compiling the specific contractors, engineering firms, carriers, counsel, and registered lobbyists named in the NCDOT encroachment filings, procurement exhibits, and Secretary of State registry entries for New Hanover County and the surrounding districts. Every entity we name here will appear on a filed document, with the document cited. Inference does not go in this section. That record will be published here as it is assembled.

Part Eight — What to Pursue

Take the cameras down.

The commissioners take this up again in September, so let’s be clear about the goal, because it is not to run this surveillance more neatly. The objective is to end the program and remove the cameras. A permanent, searchable record of where innocent people drive is not made acceptable by adding a warrant requirement, a shorter retention window, or a cleaner audit log — those administer the dragnet; they do not end it. New Hanover County did not have this network a few years ago and it functioned. It can again. Cancel Contract #25-0364, remove the readers, and stop building the county’s camera network out any further. That is the ask.

If the Board will not go that far, the items below are the floor — the minimum, not the destination. Take them as harm reduction while the real fight — removal — continues.

01 · Close the national lookup gap. Neither the Sheriff’s August 10 changes nor Flock’s August 13 changes address who outside this county may search data collected here. That single omission accounts for the overwhelming majority of the 2.98 million searches. Require an opt-out from national lookup as a condition of renewal.

02 · Require a warrant or documented individualized suspicion. Article 3D does not require one. Nothing prevents the county from requiring one by contract and by policy. A written case number, a supervisor’s approval, and a retained justification cost the Sheriff’s Office nothing when the search is legitimate.

03 · Publish the audit, unredacted as to agency. The Washington Post’s central finding was that departments review audit logs only after a complaint. Monthly internal audits are an improvement. Monthly published audits — naming the searching agency, if not the officer — are accountability. Redacting the agency defeats the purpose of the log.

04 · Bar immigration and reproductive-care enforcement in writing. Illinois wrote this restriction into statute and a state audit still found violations — because the statute carried no penalty. North Carolina’s Class 1 misdemeanor is a real penalty. Attach the restriction to the contract, name the remedy, and audit against it.

05 · Ask what happens on the state right-of-way. The budget signed July 7, 2026 lets the SBI place readers on NCDOT roads with NCDOT approval. Austin cancelled its contract and got state cameras instead. Commissioners should ask, on the record, how many state-authorized readers are approved, pending, or installed in New Hanover County.

06 · Ask who signed. Request the encroachment agreements, the bonds, the interlocal agreements, and the procurement exhibits. Not to accuse anyone. To establish, in the record, who is accountable for what sits on a public road in this county.

None of this requires anyone to accept a theory about the future of money. It requires reading a contract and deciding that a community is not a database to be searched. That is ordinary civic diligence, and this county has already demonstrated it can do it.

The convergence argument matters for a different reason. It explains why a camera on Market Street is worth this much attention when it looks, on its face, like a minor municipal purchase. The camera is the part of the surveillance state you can still see, still vote on, and still take down — before it is wired to everything else. So take it down.

Join the 3,741+ residents who’ve signed Email the commissioners →

Sources

1. WECT Investigates, Aug. 18, 2026 — “Records show nearly 3 million Flock searches in New Hanover County since 2025.” link
2. WECT Investigates, Aug. 10, 2026 — “Sheriff addresses Flock concerns, announces policy changes.” link
3. Port City Daily, Aug. 18, 2026 — “’Get the flock out of our lives’: NHC residents protest surveillance cameras.” link
4. 404 Media, May 29, 2025 — “A Texas cop searched license plate cameras nationwide for a woman who got an abortion.” link
5. EFF, Oct. 7, 2025 — “Flock Safety and a Texas sheriff claimed a license plate search was for a missing person. It wasn’t.” link
6. 404 Media, May 27, 2025 — “ICE taps into nationwide AI-enabled camera network, data shows.” link
7. NPR Illinois, Aug. 27, 2025 — “Hundreds of police departments use camera company accused of breaking state law.” link
8. 404 Media, Dec. 22, 2025 — “Flock exposed its AI-powered cameras to the internet. We tracked ourselves.” link
9. The Washington Post, Aug. 2, 2026 — “How rogue officers turned a nationwide camera network into a tool for stalking.” link
10. The Washington Post, Aug. 19, 2026 — “We found cops who misused Flock. Their police departments didn’t know.” link
11. Stateline, Aug. 14, 2026 — “Flock Safety tightens safeguards as states, cities question surveillance network.” link
12. Center for Human Rights and Privacy, Apr. 30, 2022 — “Dissection of a Flock Safety camera.” link
13. N.C. General Assembly — N.C.G.S. Chapter 20, Article 3D, §§ 20-183.30 to 20-183.33. link
14. UNC School of Government, Nov. 6, 2024 — “License plate readers: highway pilot program and new offense.” link
15. Carolina Journal, Aug. 7, 2026 — “NC budget makes license plate readers permanent amid privacy concerns.” link
16. NCDOT — Encroachment Agreements (Forms 16.1, 16.1A, 16.2). link
17. Spectrum News, Feb. 13, 2026 — “Texas DPS license plate readers in Austin draw pushback.” link
18. Supreme Court of the United States, Jun. 29, 2026 — Chatrie v. United States, No. 25-112. link
19. The Record, Jul. 10, 2026 — “License plate cameras may be next target after Supreme Court reins in location tracking.” link
20. WHRO, Feb. 11, 2026 — “A federal judge ruled Norfolk’s Flock surveillance cameras don’t invade people’s privacy — yet.” link
21. IMF, Oct. 19, 2020 — “Cross-Border Payments: A Vision for the Future” (Carstens remarks). link
22. IMF, Oct. 14, 2022 — “Central Bank Digital Currencies for Financial Inclusion” (Bo Li remarks). link
23. U.S. GPO — Pub. L. 119-101, § 1101 (CBDC prohibition through Dec. 31, 2030). link
24. The American Presidency Project — Executive Order 14178, Jan. 23, 2025. link
25. Federal Reserve, Aug. 5, 2021 — Gov. Christopher Waller, “CBDC: A Solution in Search of a Problem?” link

By Mark Spencer Williams. Figures are cited with their dates attached because they change. Every load-bearing fact here comes from a court record, a state audit, a statute, or a named news investigation; the constitutional discussion is general information about public matters, not legal advice. Corrections are welcome and will be published — mark@deflockilm.org. First published August 22, 2026.