Have I Been Flocked? How to Check If Flock Searched Your License Plate
- You can check in about ten seconds at haveibeenflocked.com. It searches Flock audit logs that agencies have released under public records law, and the site states it does not save, store or log the plate you type.
- A blank result does not clear you. The site says plainly that its dataset is incomplete, that records are often redacted, and that there can be a delay of ‘months or even years’ between a search happening and it appearing. No result means no record here — not no search.
- A hit does not mean you were investigated. It means an operator queried your plate. The site notes the user is not necessarily a police officer, and that many agencies require neither a warrant nor suspicion to run a search.
- It cannot tell you whether a camera photographed your car. That is a different question with a different answer, and the audit logs simply do not contain it.
- This is in the news because of documented abuse. A Washington Post investigation published 2 August 2026 found at least 50 officers charged with or accused of misusing Flock to stalk people; in 26 cases investigators concluded the targets were wives, girlfriends, former partners or women the officer wanted to approach. Flock announced new controls eleven days later.
You can check whether your license plate has been searched in Flock’s system in about ten seconds, at haveibeenflocked.com . Type the plate, get a result.
The harder part is reading the answer correctly. A hit does not mean you were investigated, and a blank does not mean you were not searched. Almost everything useful about this tool sits in that gap.
How do I check my plate?
Go to haveibeenflocked.com , enter your plate, and read the result. There is no account and nothing to install.
The site states it does not keep what you type: “This website does not save, store, or log the license plate numbers or search terms you enter. The search is performed and the data is immediately discarded by the application itself.”
What you are searching is not Flock’s live database — no member of the public has access to that. You are searching audit logs: the records of who searched what, which some agencies release under public records law. On 16 August 2026 the site reported holding 4,684,253 plates across 241,961,192 searches, described in its own words as the searches “we know about”.
What does it mean if my plate comes up?
It means an operator queried your plate. That is all it means.
The site’s own FAQ is careful here, and worth quoting rather than paraphrasing: a search result means an operator “queried the database for your license plate”, and “this user is not necessarily a police officer, and a search does not imply you were stopped or investigated.”
The line most people find surprising follows immediately: “many police agencies do not require a warrant or even suspicion to perform a search.” A record of your plate being run is not evidence that anyone suspected you of anything — which cuts both ways, because it also means a search needed no justification to happen.
Each record typically carries the searching agency, a date, and an operator-supplied reason. Those reasons are free text or dropdown selections entered by a person, so they range from case numbers to a single vague word.
What does it mean if nothing comes up?
Much less than you would hope. A blank is not an all-clear.
This is the part that gets lost when the tool is written up as “find out if police searched you”, and the site says so plainly:
“The dataset is incomplete; few governments provide easy access to these logs, and the records we obtain are often redacted.”
There are three separate reasons a real search can be missing from your result:
- The agency never released its logs. Coverage depends entirely on what has been obtained through public records requests and what agencies publish on transparency portals. Most have not been.
- The records were redacted. What is released is frequently partial.
- It has not arrived yet. The site warns of “a significant delay—months or even years—between when a search occurs and when it appears on this website.”
So the honest reading of an empty result is narrow: no record of a search appears in the logs this project currently holds. That is genuinely different from not having been searched, and it is the single most important sentence on this page.
Can it tell me if a camera photographed my car?
No, and this is the most common misunderstanding.
Audit logs record searches, not sightings. The FAQ is explicit: the results “don’t show when or if your vehicle passed one of Flock’s estimated 120,000+ cameras and connected devices, or what those devices returned.”
The practical translation: if you drive regularly in an area with these cameras, your plate has very likely been photographed many times, and none of that will ever show up in this tool. Being read by a camera is routine and automatic. Being searched is a person deciding to look you up. They are different events, and only the second leaves the kind of record this site can obtain.
On how many cameras there are, the sources disagree and we are not going to pick one for you. The tool’s FAQ says “estimated 120,000+ cameras and connected devices”. CNN’s 13 August report puts it at “roughly 80,000+”. The gap may be devices counted against cameras counted; either way, the number is large and growing.
Why is this suddenly everywhere?
Because of a documented pattern of abuse, and a company response eleven days later.
The Washington Post published an investigation on 2 August 2026 finding that at least 50 law enforcement officers across the US have been charged with or accused of misusing Flock’s system to stalk people. In 26 of those cases, investigators or prosecutors concluded the technology was used to watch wives, girlfriends, former partners, an ex’s new partner, or women the officer wanted to approach.
The scale within individual cases is what makes the numbers land:
| Case | What was alleged or found |
|---|---|
| A Georgia police chief | Allegedly searched a former partner’s plate, and her teenage daughter’s, about 600 times |
| A Wisconsin officer | Allegedly used Flock to check whether an ex-partner had visited an abortion clinic |
| A former Milwaukee officer | Court documents show he searched a romantic partner’s location 124 times |
| A former California deputy | Convicted on 10 charges including stalking and unauthorised access of electronic data; sentenced in February to six years |
Three of those four are allegations rather than findings, and we have marked which is which. We have not named the officers or the people they targeted.
What did Flock actually change?
On 13 August 2026 the company announced new controls. CEO Garrett Langley called the misuse “completely unacceptable”. The specifics, with their deadlines:
| Change | Detail |
|---|---|
| Audit Assistance made mandatory | Flags abnormal search behaviour and locks the user out pending administrator review. Required for all customers by the end of 2026. About a third had switched it on voluntarily |
| Case number required | Officers must enter a case number to justify each search. “Genuine emergencies” are exempt but automatically flagged. Previously optional |
| Retention cut | Standard data retention drops from 30 days to 7. Agencies can still adjust it under state or local policy |
| Evidence mode | Lets agencies preserve data beyond the retention window when it is tied to a case number |
| Sharing limits | Local jurisdictions can restrict what outside agencies may search their cameras for — for example permitting theft searches but not immigration enforcement |
What the changes do not fix
Worth being straight about, because the announcement was widely reported as a resolution:
- They are forward-looking. Nothing in them surfaces past searches to the people who were searched.
- The retention default is a default. Agencies may set it otherwise under state or local policy.
- “Genuine emergencies” is an exception written by the party being audited. Flagged for review is not the same as prevented.
- None of it gives you a right to be told. There is still no notification when your plate is run. That asymmetry — a search needs no warrant, but finding out takes a records request — is the reason a volunteer project had to exist at all.
What to do if you find a hit
- Screenshot the record, including the agency, date and stated reason.
- File a public records request with that agency for its full audit logs. The site publishes an open records guide with request language and state-by-state examples drawn from more than 500 completed requests.
- Ask for more than the search log — network sharing settings, event logs and configuration records are all worth requesting, and often more revealing than the searches alone.
- If you believe you were targeted personally, the pattern in the reported cases is repeated searches of the same plate over time. That shape, rather than any single record, is what investigators acted on.
- Check what your own state allows. Rules on ALPR retention, sharing and disclosure vary widely, and some states restrict out-of-state queries.
Sources
| Source | What it supports here |
|---|---|
| Have I Been Flocked — FAQ | Every quoted limitation: incomplete dataset, redactions, months-to-years lag, what a search result implies, no-warrant searches, the 120,000+ device estimate, and the statement that typed plates are not stored |
| Have I Been Flocked — homepage | The 4,684,253 plate and 241,961,192 search counters, read 16 August 2026 |
| Have I Been Flocked — audit logs guide | How to file a records request, and the other Flock records worth asking for |
| Washington Post: How rogue officers turned a nationwide camera network into a tool for stalking | The 2 August investigation, the 50-officer figure and the 26 cases |
| Washington Post: Flock unveils changes after Post report | The link between the investigation and the company’s response |
| CNN: Flock rolls out new police auditing and accountability controls | Each announced change and its deadline, the 30-to-7-day retention cut, the one-third voluntary adoption figure, the 124-search Milwaukee case, the ~80,000 camera figure and the CEO’s quote |
| Poynter: A Washington Post investigation prompts changes at Flock | Independent confirmation of the investigation and that it prompted the company’s response — used because the Post itself would not load for us |
| Al Jazeera: Flock adds safeguards after wave of stalking cases | Second reachable account of the 13 August changes and the reporting behind them |
| Flock Safety | The company’s own site |
Checked 16 August 2026. No affiliate links, and no payment was received for any link on this page.
How we verified this
The limitations in this article are the site’s own words, not our characterisation. We read the tool’s FAQ directly rather than relying on news write-ups, and quote it where the wording matters — on the dataset being incomplete, on the months-to-years lag, on what a search result does and does not imply, and on the site not storing what you type. A “how to check” article that omits those turns a partial record into a false all-clear.
We verified the site is real before linking it. haveibeenflocked.com, its search page and its audit-logs guide all return live pages, and an invented path on the same host returns 404 — so the test can both succeed and fail rather than passing for everything.
The counters were read on the day of writing. The homepage reported 4,684,253 plates across 241,961,192 searches on 16 August 2026. Those numbers move as more logs are released, so they are stamped rather than presented as fixed.
The two camera counts disagree and we print both. The tool’s FAQ says “estimated 120,000+ cameras and connected devices”; CNN’s 13 August report says “roughly 80,000+”. We have not found a basis for preferring one, and the difference may be devices versus cameras, so both are attributed rather than averaged or silently resolved.
The Washington Post is cited but could not be opened from here, so we corroborated it elsewhere. washingtonpost.com returns nothing to us — not a 404 on those two articles, but no response at all, including for the site’s own front page. That means our usual check cannot tell a real Post URL from an invented one, so we did not treat “it loaded” or “it didn’t” as evidence either way. Instead we confirmed the findings through two sources we could open and control-test: Poynter’s write-up of the investigation and its consequences, and Al Jazeera’s 14 August report. The two Post URLs are given because they are the original reporting and are indexed under those exact headlines, not because we opened them.
We did not name the officers or the people they targeted. The cases are public and widely reported, and several officers have been charged but not convicted. Naming accused-but-not-convicted individuals adds nothing a reader needs in order to check their own plate, and naming the people who were stalked would repeat an intrusion the article is about. We kept the roles, the figures and the outcomes, and dropped the names. Where a case has been decided we say so; where it is an allegation we say that too.
The hero image is a generic CCTV camera, not a Flock device. This article is partly about not mistaking one piece of roadside hardware for another, so the picture is decorative and should not be used to identify anything. An earlier candidate showing a named Texas parkway was rejected for exactly that reason: the equipment on it was most likely toll sensors.
We have not tested a plate. We are not in the United States and do not have a US plate to search, so this describes what the tool states it does and what its records are, not a walkthrough of our own result.
The site has a viewpoint, and that is disclosed here. It is fiscally sponsored by the Alternative Newsweekly Foundation, states it is editorially independent from it, and says it is not affiliated with Flock Safety or any government agency. Its homepage also promotes a campaign week against ALPRs. The underlying records are genuine public records; the framing around them is advocacy, and readers should know both.