Last updated:
🔦 We Mapped a Surveillance Network With Free Tools. Here's How You Can Too.
A working investigation, not a think-piece. Using only free, public sources — a crowdsourced camera map, government contracts, court filings and published audit logs — here's how to document an ALPR deployment in your own town, reproduce the abuses that made national news, and check whether police searched your own plate.
The premise: anyone claiming a surveillance network is unaccountable should be able to show it, with sources a reader can check. So we did — using nothing but free tools, several of them on this site. This is the method, step by step, reproducible for any town. By the end you will be able to map the cameras near you, confirm who paid for them, find any documented misuse, and check whether your own plate has been searched.
By Ned Walsch · a Max Intel investigation
16 Oct 2026 — corrections: the Johnson County search covered “more than 83,000” cameras, not 83,345. GainSec lists 51 security findings, not 51 CVEs. Institute for Justice counts are updated. DeFlock is now described as a map of OpenStreetMap data, and deflocktheusa.com is no longer called DeFlock’s list.
- Retention is now agency-specific, so ask before you file. Flock's 13 August 2026 safeguards cut the default retention to 7 days for new deployments while existing customers keep their current periods (30 days was the old default; Henrico VA kept images for 21 days). For step 4 that means the plate images behind a search may be gone within a week, but the audit log, case codes (mandatory for all searches by end of 2026) and Audit Assistance flags are separate records and are what to request. Flock Safety (13 Aug 2026); what to ask for is set out in the Flock guide's OS Investigate section and safeguards table.
- The contract may already be ending. The Institute for Justice database lists 277 local governments that cancelled or did not renew since the beginning of 2025 (as of 15 Oct 2026). Henrico County VA announced on 6 October that it was terminating its contract, after installing more than 120 readers over four years. Before step 2, check the IJ database, deflocktheusa.com’s cancellation list and Who Approved This for your town's vote record and renewal date; a cancelled contract makes the records request more urgent, not less, because the logs leave with the vendor. VPM (6 Oct 2026)
Key findings
A single Flock search in Johnson County, Texas, reportedly to locate a woman over an abortion, covered more than 83,000 cameras 404 Media (29 May 2025). The sheriff said her family feared for her safety. One query reached 6,809 camera networks, because agencies that share access in Flock can search each other's data EFF (30 May 2025). The Institute for Justice lists 279 incidents of ALPR abuse. Its separate stalking list holds 82 cases of police allegedly using ALPR to stalk (both read 16 October 2026). By the Institute for Justice's count, 277 local governments had cancelled or declined to renew ALPR contracts since the beginning of 2025, as of 15 October 2026. Independent researchers (GainSec) catalogued 51 security findings in Flock's own devices and systems; not all were submitted for CVE (public vulnerability ID) assignment. Every one of those facts came from public sources — procurement records, court filings, published audit logs and reporting — and this page shows you the exact five steps to reproduce the method for your own town, ending with whether your plate was searched.
Why ALPR, and why now
Automated license-plate readers do not photograph suspects. They photograph everyone — every plate that passes, timestamped and geotagged, stored and searchable. Flock Safety, the largest US vendor, operates cameras across thousands of communities, and until recently almost none of it was visible to the people being recorded.
Over the past year that changed, because the public record filled with evidence of what a lightly-audited, mass-search system does at the margins. Three facts from that record frame everything below:
The investigation, in five reproducible steps
Every step uses a free tool. Do them in order for your own town and you will have a sourced picture of the surveillance around you in an afternoon.
What the method surfaces
Run those five steps against the deployments already in the public record and a consistent shape emerges — the reason the contract terminations happened:
- Scale outruns oversight. A system searchable from any connected agency means a local camera is a national instrument. The Johnson County search covered more than 83,000 cameras in 6,809 networks because agencies share access with each other — one login, thousands of agencies' data.
- Audit trails are thin where it matters. The misuse cases share a signature: searches logged with vague or absent justification. The log exists; the check on it did not.
- The harms are not aimed at criminals. An officer allegedly tracking an ex. A search to follow a woman seeking an abortion. These are not edge-case failures of a crime-fighting tool — they are what a personal-movement database enables when someone with access decides to use it.
- Legislation followed the scrutiny. Washington's SB 6002 (signed 30 March 2026 Washington Legislature) and the wave of terminations did not happen in a vacuum. They followed exactly the kind of public documentation this method produces.
- The security is not airtight either. Independent researchers (GainSec) catalogued 51 security findings in Flock's own systems — the surveillance infrastructure is itself a target, which means your movement data is only as safe as the vendor's patching.
The honest counter-case
Accountability work is worthless if it only argues one side, so here is the strongest case for ALPR, made fairly. These systems solve real crimes — recovered stolen vehicles, AMBER Alert hits, homicide leads that would otherwise go cold. The vast majority of queries are routine and legitimate. Vendors point out that audit logs exist precisely so misuse can be caught, and that the documented cases, while real, are a tiny fraction of total searches. A town facing a genuine crime problem is not irrational to want the tool.
The rebuttal is not that ALPR never helps — it is that "mostly used well" is not the standard we apply to mass surveillance of people who are not suspected of anything. A system that records everyone, searchable millions of times, with 82 police stalking cases on the Institute for Justice’s list as of 16 October 2026, is one whose costs belong in public view alongside its benefits. That is what this method puts there. Readers can weigh the trade for their own town — which is the entire point of making it reproducible.
Frequently asked questions
Is it legal to investigate ALPR cameras like this?
Yes. Everything in this investigation uses public data: crowdsourced camera maps, public procurement records, court filings, published audit logs and news reporting. No access control is bypassed and no private system is entered — mapping cameras you can see from a public road, and reading government contracts and court documents, is squarely lawful open-source research. The line would be crossed only by accessing a camera network's back end without authorisation, which nothing here does. See our guide on whether OSINT is legal.
Can I really check if police searched my own plate?
In many jurisdictions, yes — and it is the single most concrete step in this piece. haveibeenflocked.com searches published Flock audit logs, which some agencies release, and can show whether your plate was queried, by which agency, and under what stated reason. Coverage depends on which agencies publish their logs, so absence of a result is not proof you were never searched. It is a genuine, reproducible check that turns an abstract surveillance debate into a personal one.
What is DeFlock and how accurate is it?
DeFlock (deflock.org) is a crowdsourced map of automated license-plate readers. It displays cameras that volunteers have tagged in OpenStreetMap, using the DeFlock app or OpenStreetMap editors. It is not exhaustive and it is not official, so treat it as a floor, not a census: the real number of cameras in any area is at least what DeFlock shows and usually more. Its value is that it is a public, independent map of a surveillance network the vendor does not publish.
How do I reproduce this for my own town?
Follow the five steps in the walkthrough below, all with free tools. Start with DeFlock to map local cameras, cross-reference against public procurement and city-council records to confirm the contract and cost, search news archives and court filings for any documented misuse, and finish with an audit-log check for your own plate. Each step uses a tool linked from this page. The method is the point — it works for any ALPR deployment, not just the cases we cite.
Why does this matter if I have nothing to hide?
Because ALPR does not record suspects — it records everyone, continuously, and stores where your car was and when. The documented harms in this piece were not aimed at criminals: they include an officer allegedly stalking a former partner, and a nationwide search run to track a woman over an abortion. The risk is not that you did something wrong; it is that a permanent, searchable record of your movements exists and can be queried by anyone with access, for any reason their oversight fails to catch.
Is this an anti-police investigation?
No. It is an accountability investigation, and the distinction matters. Automated license-plate readers are a legitimate tool with legitimate uses, and most queries are routine. The point is that a system searchable millions of times with weak audit trails will be misused at the margins — and the public record now shows it has been. Documenting that is not opposition to law enforcement; it is the ordinary work of open-source scrutiny that applies to any powerful system operating with limited oversight.
Related
See also: DeFlock’s ALPR map for the camera locations this walkthrough maps · License-Plate Readers and the 2026 Midterms.