Durham Calls Its Flock Records a ‘Criminal Investigation.’ Who’s Being Investigated?

What Durham said

Reporting on records requests across the Triangle, INDY Week found that the Durham Police Department has taken the position that its Flock usage logs “constitute a record of a criminal investigation … and are not public records.” (INDY Week) New Hanover County does a narrower version of the same thing: it produced its 2,980,082-search audit with the searching agency blacked out in every row, citing confidentiality. That New Hanover refusal is not secondhand: the County Attorney’s written rationale sits in our records hub, in the County’s own words. Durham’s blanket “criminal investigation” position is the sharper national version, and the actual denial letter is a public record worth prying loose.

Then who is the suspect?

Start with the question the label forces, because it is the whole story. A criminal investigation, as everyone understands the phrase, has a subject — a person the police suspect of something. Durham’s Flock cameras captured the plates of everyone who drove past them: commuters, churchgoers, delivery drivers, teenagers, retirees. If the record of all of that is “a criminal investigation,” then when, exactly, did all of those ordinary drivers become the subjects of one?

This is where automated plate reading inverts the order the Fourth Amendment assumes. Traditionally, police start with a suspect, develop cause, and then investigate. Flock runs the sequence backward. It collects everyone first, stores everyone first, and makes everyone searchable first — and only later decides who turns out to be relevant. Put plainly: traditionally, police investigate suspects. This investigates databases. And the databases contain everybody.

Make it concrete, because that is where surveillance stops being abstract. The nurse driving to a 6 a.m. shift. The family pulling into church on Sunday. The parent taking a child to a counselor’s office. The resident driving downtown to speak against the cameras at a commission meeting. Every one of those plates is captured, time-stamped, stored, and dropped into a system a police department is willing to call — in writing, to fend off a public-records request — a criminal investigation.

It also answers the oldest line in the surveillance debate. Told “if you’ve done nothing wrong, you have nothing to hide,” a resident now has a straight reply: then explain why the record of my ordinary commute is sitting inside something the government itself describes as a criminal investigation. That uncomfortable question is not ours. Durham raised it.

The description that cuts against the County — as evidence, not a gotcha

Now the law, carefully, because the honest version of this argument is stronger than the overstated one. Whether a record must be disclosed under the public-records law and whether collecting the data was a Fourth Amendment “search” are two different questions, and one does not automatically answer the other. A record can be exempt from disclosure while the same collection is found not to be a search. So we are not claiming a records label creates a constitutional violation.

What we are claiming is narrower. When agencies defend warrantless plate readers in court, the argument is that this is routine, limited observation: a plate on a public road is exposed to anyone, and dispersed cameras leave gaps that fall short of tracking the whole of a person’s movements — the rationale in United States v. Martin (E.D. Va. 2024) and the Norfolk case below. Then a resident asks for the records, and the same data is described as a criminal investigation. Those two descriptions are in tension, and the County’s own characterization of its data as investigative is evidence a resident can use to challenge the claim that the system is merely routine, limited observation. It doesn’t settle the constitutional question. It bears on it.

What § 132-1.4 actually requires

The exemption the County leans on is narrow and specific. N.C. Gen. Stat. § 132-1.4 removes from the public record “records of criminal investigations” — defined as records or information “that pertains to a person or group of persons” and is “compiled by public law enforcement agencies for the purpose of criminal investigation.” And the same statute contains a line that matters here: “The use of a public record in connection with a criminal investigation … shall not affect its status as a public record.” An administrative audit log does not become secret merely because searches recorded in it were investigative.

So a categorical “it’s all a criminal investigation” does not answer the questions the statute actually asks. A resident — or a court — is entitled to make the County be specific: Which field is withheld? Does it “pertain to a person or group of persons”? Was the audit log “compiled for the purpose of criminal investigation,” or is it an administrative record of how the system was operated? Is the County claiming criminal-investigation records, criminal-intelligence information, or some other exemption? A blanket assertion answers none of them. North Carolina’s courts recognize both the broad public-access baseline and this narrow exemption; the agency asserting it has to fit the record to the statute. (Gannett Pacific Corp. v. N.C. State Bureau of Investigation, 164 N.C. App. 154 (2004).)

An audit log is not a case file

Press on the distinction, because it is where the County’s position is weakest. A Flock audit log records how the system was operated — who logged in, what was queried, when, and by which agency. That is operational metadata, generated automatically by the platform; it is not evidence gathered in the course of investigating a particular crime. It is the administrative record of a government program — the same “who accessed this, and when” log the Sheriff’s own policy already requires be kept for oversight. Calling an entire access-and-query history a “criminal investigation” treats a system’s operating record as if it were the case file. They are not the same thing, and the statute does not let an agency merge them by relabeling.

What the County will say — and the answer

Anticipate the defense, because it has a kernel of truth. The County may respond that the log does contain investigative material: hit alerts on wanted vehicles, queries tied to active warrants, stolen-car lookups connected to open cases. Some individual entries may qualify. But North Carolina law already handles that: § 132-6(c) forbids withholding an entire record because confidential information is mixed in — the agency must separate the exempt material and produce the rest. So the County can redact the genuinely investigative lines and still release the administrative backbone: the agency names, the search counts, the access metadata. “Some of it touches active cases” is a reason to redact those rows, not to seal 2,980,082 of them. And if the County refuses to segregate at all, the question only sharpens: is it protecting active investigations, or the fact that almost none of these searches were ever about one? (How to press a segregability demand.)

The Fourth Amendment question, where it really stands

The constitutional question is separate, and genuinely unsettled. In Carpenter v. United States (2018), the Supreme Court held that obtaining 127 days of a person’s cell-site location records was a search, recognizing a reasonable expectation of privacy in the whole of one’s movements.

The authority that matters most in North Carolina is binding here: in Leaders of a Beautiful Struggle v. Baltimore Police Department (4th Cir. 2021, en banc), the Fourth Circuit — the court whose federal-constitutional rulings govern North Carolina — held that Baltimore’s aerial-surveillance program implicated the Fourth Amendment because it let police reconstruct the whole of people’s movements from a retrospective, citywide database. That program was sweeping (roughly 90% of the city in daylight, 45-day retention), so it marks the direction of the law, not a decision on facts like ours.

And here is the adverse authority, stated plainly, because honest advocacy names it. On January 27, 2026, a federal district court granted summary judgment for the city in Schmidt v. City of Norfolk (E.D. Va.; our breakdown of the ruling), holding that Norfolk’s system — 176 cameras in about 75 clusters, generally retaining data around 21 days — left enough temporal and geographic gaps that it did not reveal the whole of a person’s movements, even for plaintiffs captured hundreds of times over months. Martin held the same about three captures of one vehicle over 30 days. Both are federal district-court decisions from Virginia, not binding on North Carolina courts — but they are current, they cut against the argument, and pretending otherwise is exactly the overstatement this movement can’t afford.

What would actually decide it: New Hanover’s own facts

Whether New Hanover’s network looks more like Carpenter and Leaders or more like Martin and Schmidt turns on facts, not the word “investigation.” How many cameras, and where — do they cluster at the ways in and out of neighborhoods? What is the retention period, and are downloaded records kept longer? Can agencies outside the county reach the data? How many searches run per day, per agency, per vehicle? Can “vehicle fingerprint” and route tools reconstruct a person’s recurring trips, home, workplace, or visits to a clinic, a church, a lawyer, or a protest?

Those are precisely the facts the County is withholding — which drops the constitutional question and the records question back into the same room.

The general-warrant echo

There is a reason this unsettles people across the political spectrum, and it is worth stating as what it is: a normative argument, not a court holding. A criminal investigation aimed at no one in particular and everyone at once, retained and searchable after the fact, is the kind of thing many will see as resembling the evil the Fourth Amendment’s particularity requirement was written to prevent — the general warrant, which let officers search broadly without naming a person or a cause. No court has labeled this ALPR program a general warrant, and we don’t claim one has. But the resemblance is why the doctrine keeps moving, and why a system a police department itself calls “investigative” deserves harder questions than it has gotten.

The honest bottom line

Durham’s position raises a simple question with no comfortable answer: if records generated by ordinary drivers are part of a criminal investigation, when did ordinary drivers become subjects of one? New Hanover should not be allowed to invoke the criminal-investigation exemption categorically without explaining why the particular audit-log information satisfies § 132-1.4 — which field, pertaining to whom, compiled for what purpose. And the County’s own description of the data as investigative is relevant evidence in the debate over whether its system’s scale, retention, and query power exceed ordinary public observation.

So the fair demand for August 17 is not a courtroom verdict. It is that the County pick a lane and defend it. If the audit log is an ordinary administrative record, produce it — the statute says using it in investigations doesn’t make it secret. If the County insists it is investigative, it should say which field, about whom, and compiled for what — and then explain how a system it calls investigative, run against every driver, is also the routine, limited observation it tells the courts it is. The County is 97 days into withholding the one field that would let anyone check. If you are ever told your own records are a criminal investigation, here is how to answer.

Go deeper on the constitutional questions. We have written the Fourth Amendment argument out in full, under both the federal and the North Carolina constitutions: the cases that will decide whether this is legal, the Norfolk lawsuit, what Chatrie means for Flock, why “no privacy in public” does not settle it, and — for lawyers — our Flock/ALPR litigation resources, covering both the federal and the North Carolina constitutions.

See the full record of what we asked and what the County withheld in the accountability clock, and — if you want the commissioners to answer the question — here is how to speak on August 17.

This article is legal and political commentary on a matter of public concern — not legal advice, and not a prediction of how any court would rule. It does not assert that any agency violated the law by responding to a public-records request. Carpenter and Leaders of a Beautiful Struggle are federal precedent (Leaders binds North Carolina; its facts are far more pervasive than the ALPR record to date). Schmidt v. City of Norfolk and United States v. Martin are federal district-court decisions from Virginia and are not binding on North Carolina courts. Durham’s records position is drawn from the reporting linked above; New Hanover’s redaction is documented in our records hub. If any characterization here is inaccurate, write to mark@deflockilm.org and we will correct it in public. First published August 3, 2026.