Flock Cut Retention to Seven Days. That Is a Discovery Problem, Not Just a Privacy Story.

On August 13, 2026, Flock Safety announced it was cutting its recommended default retention period for automated license plate reader data from 30 days to seven. Most coverage treated it as a privacy concession, offered while the company was losing contracts and absorbing criticism from civil liberties groups.

For a defense team, it may be worth reading differently. Seven days is a short window, and in many cases it closes before the client has retained counsel.

What actually changed

Flock’s own announcement is worth reading directly. The company set its original default at 30 days nine years ago. It now recommends seven, reasoning that more than 90 percent of searches conducted without a full plate number happen within a week.

A few details seem more consequential than the headline number.

It is a recommendation, not a switch. Flock has said existing customers keep whatever retention period their jurisdiction already approved. Some agencies remain at 30 days. Some are shorter, some longer. The agency in any given case may or may not have adopted the new default, and it seems risky to assume either way.

Flock’s published policy already says seven days. The company’s License Plate Reader Policy, last updated June 30, 2026, describes LPR data as hard deleted on a rolling seven-day basis by default, adjustable where a customer’s law or policy requires something else. The public document and the individual agency’s contract are different things. The second one is the one that matters.

Evidence Mode runs in one direction. Alongside the retention change, Flock introduced a feature that lets detectives flag specific ALPR data for preservation in cold storage for an active investigation. That is a real improvement in one sense. It is also controlled entirely by the investigating agency. The state decides what to freeze. Nothing in the design contemplates a defendant, or a defense investigator, identifying reads that help the defense and asking for those to be held.

That asymmetry is worth sitting with. The prosecution can preserve the read that places a client near the scene. Nobody is positioned to preserve the read that places him twelve miles away forty minutes earlier, or the read showing a similar vehicle with a different plate passing the same camera.

What North Carolina law allows

North Carolina has regulated ALPR systems since 2015 under Article 3D of Chapter 20. Three provisions seem especially relevant here.

The 90-day cap and who can extend it. G.S. 20-183.32(a) says captured plate data may not be preserved for more than 90 days. Subsection (b) allows longer preservation through a search warrant or a statutory preservation request. Subsection (c) describes that request: it is made “upon the request of a law enforcement agency,” in a written, sworn statement identifying the cameras, the plate, the dates, the case and parties, and specific articulable facts connecting the data to an ongoing investigation.

Article 3D’s definition of “law enforcement agency” includes prosecutors. It does not include the defense. Whether that omission was deliberate or simply not considered, the practical effect is the same: ==the only parties with statutory authority to freeze ALPR data are the ones prosecuting the case.==

The data is confidential and not a public record. Under subsection (e), captured plate data may be disclosed only to criminal justice officers for a legitimate law enforcement purpose. A public records request will not reach it. A subpoena to the vendor runs into Flock’s consistent position that the agency, not Flock, owns the data.

Every agency has a written policy, and it covers retention. G.S. 20-183.31(a) requires an agency to adopt a written ALPR policy before the system goes live, and that policy must address data retention, sharing, training, oversight, and access. Subsection (c) separately requires agencies to keep maintenance and calibration records on file. Those documents exist for every North Carolina agency running ALPR, and they may be worth asking about early.

Taken together, the route in North Carolina appears narrow: the defense can ask the prosecutor and the agency to preserve, and can ask the court to order it, but cannot invoke the statute directly. ==That makes the timing and documentation of the request matter more than they otherwise would.==

The discovery hooks

The state’s discovery obligation under G.S. 15A-903(a)(1) extends to the complete files of all law enforcement and investigatory agencies involved in the case. ALPR queries, alert records, and audit logs generated in the course of an investigation would seem to fall within that file. A written discovery request that names them specifically, rather than relying on general categories, may be worth considering.

If records are not produced, G.S. 15A-910 gives the trial court a range of sanctions, from ordering production to dismissal. Two features of that statute are worth noting. The court is not required to impose any sanction, and before doing so it must weigh the materiality of the subject matter and the totality of the circumstances. Trial courts have wide discretion here, and appellate courts rarely disturb it. ==A record showing an early, specific request and a documented non-response tends to give the court more to work with than a complaint raised for the first time close to trial.==

How North Carolina treats evidence that no longer exists

When the question shifts from “not produced” to “destroyed,” the analysis changes.

The North Carolina Supreme Court in State v. Taylor, 362 N.C. 514 (2008), drew a distinction that has governed since. The state’s failure to preserve material, exculpatory evidence is analyzed under Brady, and no showing of bad faith is required. The failure to preserve evidence that is merely potentially useful is analyzed under Arizona v. Youngblood, 488 U.S. 51 (1988), and the defendant must show the state acted in bad faith.

State v. Williams, 362 N.C. 628 (2008), decided the same day, shows what the first category looks like. Prosecutors created and then destroyed a poster that was favorable to the defendant and usable to impeach a state witness. The court affirmed dismissal under G.S. 15A-954(a)(4) as a flagrant constitutional violation causing irreparable prejudice, without any bad-faith inquiry.

The difficulty is that Williams is unusual. The contents of the poster were known and undisputed. ALPR data that has already been purged is, almost by definition, evidence whose contents are unknown. That tends to push it into the Youngblood category, where North Carolina’s appellate courts have repeatedly declined to find a due process violation absent bad faith. State v. Hunt, 345 N.C. 720 (1997), State v. Mlo, 335 N.C. 353 (1994), and State v. Banks, 125 N.C. App. 681 (1997), are representative: negligent or even rule-violating destruction, without evidence of improper motive, was not enough.

This is the reason early action seems to matter so much. ==A specific, written, dated preservation request, made while the data still exists and ignored, is the kind of fact that can move a court’s view of good faith.== The same request made after the purge cannot.

A note on an out-of-state case. State v. Simonson, No. 40732-2-III (Wash. Ct. App. Jan. 29, 2026), is unpublished and has no authority in North Carolina. It is mentioned here only because its timeline is a clean illustration of the arithmetic. A Flock hit on June 15 led to an arrest that morning. The charge was filed June 21. Defense counsel appeared June 25. By the time counsel asked for the Flock data in early August, the 30-day window had closed and the images were gone; audit logs, kept indefinitely, were requested and never produced. Under a seven-day default, the purge would have occurred around June 22, one day after charging and three days before the defense entered the case. The Washington appellate court ultimately reversed a dismissal on state-procedure grounds that do not translate here. The calendar does.

Questions worth asking in the first week

None of this is a checklist, and the right approach varies with the case and the client. These are the questions that seem most useful to raise early when ALPR may be involved.

• Which agencies may have queried the system? Flock’s network model means neighboring jurisdictions, and sometimes federal agencies, may have searched the same cameras. The arresting agency is a starting point, not the whole picture.
• What is the actual retention period for this agency’s account? The agency’s written policy under G.S. 20-183.31 should say. Its Flock Transparency Portal, if it has published one, may say as well, and the two sometimes differ.
• Was anything archived or placed in Evidence Mode, and by whom? If the state has preserved reads, it may help to know which ones, and to ask whether others from the same window were considered.
• Is a written preservation request to the prosecutor and agency custodian appropriate now? If so, naming the plate, the known camera locations, and a time window wider than the alleged offense may be worth considering, along with keeping proof of service.
• Are maintenance and calibration records available? G.S. 20-183.31(c) requires them to be kept on file. They are the natural foundation for any accuracy question.

A sample preservation request

For attorneys who want a starting point, we’ve prepared a sample preservation request letter adapted to North Carolina. It is addressed jointly to the District Attorney and the agency records custodian, since under G.S. 20-183.32(c) those are the parties who can invoke the statutory preservation mechanism, and it enumerates the categories of ALPR material drawn from Flock’s own published policy and from Article 3D.

Download the sample letter (PDF)

The first page contains notes for counsel and is meant to be removed before sending. The letter is general information rather than legal advice; citations should be verified and the content adapted to the case, the agency, and the vendor involved.

Two other resources may also be useful. The ACLU of Massachusetts’s Surveillance Discovery Project offers model discovery and preservation motions for ALPR and other surveillance technologies; access is limited to verified defense attorneys and is requested by email through that page. NACDL’s ALPR Primer is public and outlines the discovery categories worth covering, and its Fourth Amendment Center offers direct litigation assistance.

What the records actually contain

Flock’s LPR Policy enumerates what a single LPR record includes: license plate image, vehicle image, vehicle characteristics such as color and make, plate number, plate state, date, time, and camera location. A screenshot of the hit is not the record.

The same policy describes a separate audit record for every query: username, date, time, purpose of the query, and the plate or other elements used to search. That record lives on a different schedule from the images and may persist long after they are gone. It seems worth requesting by name.

Beyond the single read the state has chosen to rely on, other items that may be useful to ask about include all reads of the subject plate across the network for the relevant window, the confidence score attached to each read, whether the result came from an automated alert or a manual or partial-plate search, which hotlist or database the plate was matched against and when it was last updated, and every query touching the plate by any agency, with the stated purpose for each.

Where this runs out

None of this makes ALPR data easy to obtain. Most courts to consider the question have found no reasonable expectation of privacy in a plate displayed on a public road, so suppression motions built on the Fourth Amendment alone have often struggled. The preservation and discovery route has its own limits, described above.

What the retention change alters is the calendar. When the window was 30 days, a defense team retained in the second week still had a reasonable chance of finding the data intact. At seven days, it will frequently be gone before arraignment. That does not dictate any particular strategy. It does suggest that the question of ALPR preservation is worth raising earlier than it used to be, and in writing.

This post is general information for defense teams and is not legal advice. Retention rules, preservation mechanisms, and discovery obligations vary by jurisdiction and by agency. The law and policy applicable to a specific case should be confirmed independently.