Leadership1 distinct publisher3 min readUpdated
The camera company has 120,000 license plate readers in 49 states and 7,000 agencies under contract. Dozens of cities and counties have now cut ties, and the hardware is being vandalised.
The Board Room · Leadership desk

Compiled by The Board RoomSomething wrong?How this is made
Flock has put 120,000 automated licence plate readers on roads in every US state except Alaska, and 7,000 law enforcement agencies now use them, roughly 40 percent of the nation's police departments [1][4]. Dozens of cities and counties have cut ties with the company, its cameras have become targets for vigilante vandalism, and the argument is being had at council meetings rather than in Flock's boardroom [5][6][7].
That is the structural feature worth studying. Flock, founded in 2017 and most recently valued at $8.4 billion, grew by selling to agencies, not to electorates [2][3]. Its cameras scan the plate of every passing car along with make, model, bumper stickers and other identifying features, and what distinguishes the product is an AI-powered national database that police can search with very few restrictions [8][9]. The 40 percent figure implies a total addressable base of about 17,500 departments, and an average of roughly 17 cameras per customer agency [10][11]. The unit sales are small; the network effect is the asset. So is the exposure.
The utility case is made by the buyers. Flock says its cameras have been used in a million criminal investigations a year, and police say cross-jurisdiction tracking has been critical to solving thousands of crimes including car thefts, missing persons cases and child abductions [12][13].
The failure case is also made by the buyers, which is the problem. According to 404 Media's Jason Koebler, who has reported on the company for years and spoke to Vox's Today, Explained, local police helped ICE surveil undocumented immigrants through the Flock network, with officers in Texas, South Carolina and Tennessee searching nationwide using "immigration" or "ICE" in the system's reason field [14][15]. Those searches reached Illinois and California, where using the technology for immigration enforcement is against state law, and an Illinois secretary of state investigation found the searches broke state law, after which Flock made changes [16][17]. In a Texas case, after a woman's partner reported she had self-administered an abortion, more than 80,000 cameras across the country were searched for her, about two thirds of the entire deployed network [18][19]. She was not located, but police there considered whether they could charge her [20]. Separately, there are dozens of reported cases of officers illegally searching Flock to track current or former romantic partners [21].
Note how those came to light: public records requests filed by journalists and citizens, not by any control inside the system [22]. The reforms Flock has announced are dismissed by critics as window dressing, and have not slowed the cancellations [23][24]. When the podcast team asked Flock for comment, it received an out-of-office reply saying its media team was "currently touching grass and taking a break" [25]. Vox pulled one line from the interview to summarise the position: "There's been a political backlash to it, and there's also been a vigilante backlash to it" [26].
The lesson for anyone selling infrastructure into public agencies is that consent cannot be procured wholesale. The chief who signed a contract for 17 cameras did not sign up for a searchable national fleet of 120,000, and now owns the consequences locally [11][1].
Watch the cancellation count, and whether departures cluster in states with statutory limits on the data's use, as Illinois and California have [5][16]. Watch whether disclosure of misuse continues to arrive by records request rather than from Flock's own audit machinery [22]. And watch what a network worth $8.4 billion is worth if a meaningful share of its 7,000 agencies exit [3][4].
Follow any of these and your For You feed starts watching them — no settings page required.
Ranked by verification strength, evidence, and original report placement.
Flock has put 120,000 cameras on roads in every US state except Alaska.
7,000 law enforcement agencies across the country use Flock cameras, representing 40 percent of the nation's police departments.
Dozens of cities and counties across the country have cut ties with Flock.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Detailed but single-publisher and vendor-silent
Everything in the cluster comes from one Vox podcast write-up. Within it, provenance is uneven but not weak: the misuse accounts are attributed to a named reporter with a multi-year beat and are anchored by an Illinois secretary of state finding and public records requests, while the scale, valuation and utility figures are company or Vox characterizations with no document, filing or dataset in the cluster. Flock offered no substantive response, and the reforms at the center of the dispute are never itemized, so the most contested element is the least documented.
Near-saturation US law enforcement footprint
Adoption is the best-evidenced dimension: 120,000 cameras in 49 states, 7,000 agency customers described as 40 percent of US police departments, an $8.4 billion valuation, and a vendor claim of a million investigations a year. The countervailing signal -- dozens of cancelling cities and counties -- is real but small and unsized against roughly 7,000 agencies, so it does not yet register as measurable churn.
Backlash framing outruns the numbers on both sides
Slightly overstated overall. The documented abuse patterns are concrete and specific, so the privacy critique is not inflated; but the story's central proposition -- that the bill is arriving and the company's survival is in question -- rests on 'dozens' of unnamed cancellations and 'widespread outrage' with no churn, renewal or revenue figures against a 7,000-agency base. The vendor side is equally unverified: a million investigations a year and reforms said to prevent misuse are asserted, and the reforms are not described. The gap is therefore modest and cuts in both directions, mostly reflecting unquantified consequences rather than exaggerated facts.
Visible commercial, editorial and political stakes
Incentives are legible from the cluster itself. Flock is an $8.4 billion venture-scale company whose growth depends on municipal renewals, and it declined to speak, so its reform announcement functions as reputational defense in a procurement fight. The reporting flows through a journalist whose outlet has built a franchise on Flock misuse scoops, published inside a Vox podcast promotion that repeatedly directs readers to the full episode. Municipal officials facing angry council meetings have their own reason to cancel visibly. None of these incentives undercut the documented findings, but they shape emphasis.
Directionally solid, quantitatively thin
Confidence is moderate. The qualitative picture -- a very large deployed ALPR network, documented cross-jurisdiction misuse, a state legal finding, political and physical backlash -- is coherent and specifically sourced. What cannot be relied on from this cluster is magnitude: cancellation counts, contract exposure, the verified utility figure, and the content of the reforms. With one publisher and no vendor response, individual numbers should be treated as attributed rather than established.
security
NPS Bought Flock. Internal Paper Says License Plate Reader; Public Statement Says Traffic Counts1 distinct publisher
security
A cop, a hot list, and a query log: where ALPR actually fails1 distinct publisher
invest
A Connecticut judge just priced prompt injection: no fine, no e-filing2 distinct publishers
product
Cinemas, classrooms and ICE: smart glasses now need a venue-policy contingency1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
1 article · August 20, 2026