A decorative Flock Safety camera appears in the top-right of the banner, as if watching Crossville landmarks below.

Vendor dossier

Who is Flock — and why Axon is not the escape hatch

What the product actually does: FreeForm AI search, patent claims versus marketing, recycled press talking points, open Condor livestreams, cybersecurity findings — and why swapping to Axon is not a privacy fix.

Main Street photo: Brian Stansberry / Wikimedia Commons (CC BY 4.0) Corner Flock camera: cutout from Malwarebytes (horizontally flipped; sky removed)

Who is Flock Safety?

Flock Safety is a private, venture-backed company that sells a subscription surveillance service to police, cities, businesses, and HOAs — not a one-time camera purchase. Typical Flock contracts keep ownership of the hardware with Flock; the city pays recurring fees for installed cameras plus cloud search, alerts, and network access. Public debate is not “anti-law enforcement” — it is residents asking how a private vendor’s database is governed, audited, shared, retained, and paid for with tax dollars.

How the product actually works

Cameras capture an image of every passing vehicle plus time and location metadata. Data uploads to a private cloud dashboard. Agencies subscribe for search, alert, and analytics tools. Standard Flock service agreements state that Flock Hardware remains Flock’s exclusive property — title does not pass to the city — and Flock can remove the devices if the subscription ends. The company makes money from storing, searching, and sharing the data — not from selling the city a camera it owns.

Beyond plate lookups, Flock markets FreeForm: natural-language AI search across LPR and video (“blue SUV with a racing stripe,” “man in blue shirt and cowboy hat”). That is a searchable appearance-and-movement database, not a dumb camera on a pole.

Plain-language definition when people say “Flock sells your data”: Flock sells access to a platform built from public-space captures, plus optional analytics. Even if the vendor frames outputs as “not personal data,” it is still a paid movement database paid for by taxpayers.

Why Crossville residents should care

  • Crossville PD runs Flock Falcons per Quote Q-00234 and Atlas of Surveillance (11 cameras, Sept. 2025).
  • Scans of ordinary drivers can enter the same vendor cloud used elsewhere in Tennessee — including departments with documented personal-use misuse (Oak Ridge).
  • Contract language in other cities has included broad vendor rights to use “Aggregated Data” for marketing and product purposes (see e.g. the public Whitestown, IN Flock services agreement §4.4). Crossville should publish its live agreement so residents can read the local terms.
  • Independent reporting has also documented Flock using overseas gig workers to annotate U.S. camera imagery for AI training — see overseas AI annotation.

Overseas AI annotation of U.S. camera footage

In December 2025, 404 Media reported that Flock used Upwork gig workers — including workers in the Philippines — to review and classify vehicle and people imagery from U.S. cameras to train machine-learning models. Exposed training materials showed annotation tasks for plates, vehicle attributes, and related footage; after 404 Media sought comment, the exposed panel disappeared per contemporaneous coverage.

That is a vendor-trust and data-handling issue for any city on the network: Crossville should publish where local imagery may be processed, under what contracts, and whether council approved overseas human review. Secondary summary: DeFlock Birmingham — The Facts · The Verge.

FreeForm: Google-style search over the camera network

Flock FreeForm™ is the company’s AI natural-language search product. Instead of only typing a plate, an officer can type a description and pull visual matches from LPR and video — across the agency’s cameras and, where sharing is enabled, partner networks. Crossville residents should treat “just license plate cameras” as incomplete marketing until the city publishes whether FreeForm / people-search / video-search features are enabled on the local contract — ask under ask before renewal.

  • Vehicles without a plate: plain-language prompts such as “blue SUV with a racing stripe” or “white F-150 with a ladder in the back.”
  • People by appearance: Flock’s own marketing example is “man in blue shirt and cowboy hat.” A company spokesperson has confirmed FreeForm can search by clothing descriptions.
  • Cross-agency reach + alerts: search shared cameras where authorized; get notified when a match reappears — without watching every feed.
  • Flock’s framing: they say FreeForm is not facial recognition — clothing / accessories / vehicle attributes only. Critics note appearance-based search still enables warrantless tracking of people and cars by look across a networked map. Separately, Flock’s own U.S. patent specification describes broader people-classification embodiments — see patent vs marketing.

“Just type what you’re looking for, like ‘man in blue shirt and cowboy hat,’ and get visual matches instantly.”

Flock Safety FreeForm product page

Questions for renewal: Is FreeForm (or any natural-language / people-description search) enabled? On which camera types? Who can run those searches? Are clothing/appearance queries logged and audited? Has council approved people-search capabilities separately from basic LPR?

Sources: Flock FreeForm product page · Flock blog (claims no facial recognition) · Police1 — FreeForm launch coverage · AOL — spokesperson on clothing search · InvestigateTV — training vs. “don’t track people”

What Flock told the Patent Office vs. what it tells cities

Independent researchers at DeFlock SC flagged a tension worth checking against primary documents: Flock’s public FAQ says LPR products do not use facial recognition and do not identify race, gender, or ethnicity — while a Flock Group Inc. patent describes people-classification and face-matching embodiments that go much further. A patent does not prove that every claimed embodiment is live on Crossville’s cameras today. It does show what the company put under oath to the U.S. Patent Office to protect as IP — and what cities should demand in writing before renewal.

U.S. Patent US11416545B1 (“System and method for object based query of video content captured by a dynamic surveillance network”) lists inventors Garrett Langley and Matt Feury, assignee Flock Group Inc., filing date Oct. 4, 2020, and grant / publication date Aug. 16, 2022.

  • Patent specification (detailed description): describes neural networks that classify people into classes including “male, female, race, etc.,” plus clothing types and estimated height/weight. A worked example says investigators may obtain “face recognition data points” from one camera and query other cameras for matching face recognition data points. Object fingerprints may be treated as the same object when a similarity probability exceeds a threshold (cross-camera matching without a name).
  • Formal numbered claims: the granted claims speak more broadly about classifying “object class” and “object aspect” data and searching on those fields — they do not spell out “race” or “facial recognition” as claim limitations. Treat race / face language as what the specification discloses, not as verbatim claim text.
  • Public marketing (contradiction): Flock’s “Are Flock Products Discriminatory?” page says LPR cameras “do not use facial recognition,” “do not collect demographic information,” and “do not identify race, gender, or ethnicity.” Its rights & safeguards / civil-liberties page repeats that ALPR “does not use facial recognition” and “does not collect or infer race, religion, gender, ethnicity, or other personal traits.” FreeForm is marketed as search “without facial recognition” while still offering people-by-clothing examples (FreeForm section).
  • Already shipping (narrower than the patent’s people demographics): FreeForm’s public product page advertises natural-language people descriptions (clothing) and vehicle attributes across video/LPR — so appearance search is not hypothetical. Demographic classifiers described in the patent are a separate question cities should put in writing.
  • Contract upgrade rights in other cities: Public Flock services agreements commonly let Flock, “in its sole discretion,” make “any upgrades to [the] system or platform that it deems necessary or useful,” stating such upgrades “will not materially change any terms or conditions” of the agreement (example: Glendale, WI Flock agreement §2.12). Crossville has not published whether its live contract contains the same clause — ask under ask before renewal / TPRA.

Questions for renewal: Has anyone on council read US11416545B1? Which people-search / demographic / face-related features are contractually enabled vs. disabled? Can Flock push software upgrades without a new council vote? Will the city publish a written ban on race/gender/ethnicity classification and facial recognition if those features exist in the vendor stack?

Sources: Google Patents — US11416545B1 · Flock — Are Flock Products Discriminatory? · Flock — Facial recognition & civil liberties policy · Flock FreeForm product page · Glendale, WI — Flock services agreement · DeFlock SC — secondary analysis (Mar. 27, 2026) · full bibliography

Flock’s PR playbook: paid wires, police press-release templates, and sponsored ads

When Crossville first heard about Flock in 2022, the pitch leaned hard on Amber Alerts and finding missing people. That safety framing is part of a bigger nationwide pattern: Flock pushes company talking points into public view through paid publicity, police press kits that agencies have published in substantially similar form, and labeled ads in police trade media. Here are three documented ways that happens. (We did not find clear evidence that Flock pays TikTok or Instagram influencers for these stories.)

  1. Flock pays to blast “missing / abducted child” press releases onto the internet. The company sends its own announcements through GlobeNewswire — a service companies pay so their press releases get picked up by news aggregators. Those write-ups then show up on sites like Yahoo Finance looking a lot like regular news. Examples: Flock’s 2021 AMBER Alert / missing-children partnership announcement and a 2026 release about six abducted children recovered in Colorado with Flock plate-reader cameras.
  2. EFF and Techdirt report that Flock provided police PIO templates; several agencies published substantially similar language. The Electronic Frontier Foundation (EFF) and Techdirt report that Flock provides departments a Public Information Officer Toolkit — templates for the police spokesperson (the PIO). One documented template literally reads: “The ___ Police Department has solved [CRIME] with the help of their Flock Safety camera system. Flock Safety ALPR cameras help law enforcement investigate crime by providing objective evidence. … ____ Police installed Flock cameras on [DATE] to solve and reduce crime in [CITY].” EFF criticizes this practice as vendor-driven police messaging when that language shows up as official copy.
    Near word-for-word national examples

    This is not just a similar tone. Official pages and police press releases reuse Flock’s signature lines almost unchanged. Compare left (vendor / PIO kit language documented by EFF / Techdirt, plus Flock FAQ blocks that show up city after city) to right (what the agency published).

    Campbell, CA — official city Flock page

    Flock’s PIO kit What Campbell published
    “Flock Safety ALPR cameras help law enforcement investigate crime by providing objective evidence.” “Flock Safety ALPR cameras help law enforcement investigate crime by providing objective evidence.”
    “____ Police installed Flock cameras on [DATE] to solve and reduce crime in [CITY].” “The Campbell Police Department has installed 22 automated license plate reading (ALPR) cameras in strategic areas around Campbell to solve and reduce crime.”

    Florence, OR — police press release (Apr. 9, 2024)

    Flock’s PIO kit / vendor FAQ language What Florence PD published
    “____ Police installed Flock cameras on [DATE] to solve and reduce crime in [CITY].” “…acquisition of one automated license plate reading (ALPR) camera to solve and reduce crime. The ALPR camera was installed on April 2, 2024…”
    “Flock Safety ALPR cameras help law enforcement investigate crime by providing objective evidence.” “This system helps law enforcement solve crime by providing the objective evidence needed for investigations.”
    Vendor FAQ stock line: cameras “capture license plates and vehicle characteristics, not people or faces.” “The camera will capture license plates and vehicles characteristics, not people or faces.”
    Vendor FAQ stock line: “not intended for minor traffic or parking violations.” “…not intended for minor traffic or parking violations.”

    Foothill–De Anza campus police (CA) — ALPR page

    Flock’s PIO kit / vendor FAQ language What Foothill–De Anza published
    “____ Police installed Flock cameras on [DATE] to solve and reduce crime in [CITY].” “In February 2023, the Foothill-De Anza Community College District Police Department installed three ALPR cameras at the Foothill campus to solve and reduce crime.”
    “Flock Safety ALPR cameras help law enforcement investigate crime by providing objective evidence.” “The ALPR system also helps law enforcement solve crime by providing the objective evidence needed for investigations.”
    Vendor FAQ stock line: “capture license plates and vehicle characteristics, not people or faces.” “ALPR cameras capture license plates and vehicle characteristics, not people or faces.”
    Vendor FAQ stock line: “not intended for minor traffic or parking violations.” “…not intended for minor traffic or parking violations.”
    Vendor marketing claim repeated nationwide: “crime reductions of up to 70 percent.” “Communities using Flock Safety ALPR have reported crime reductions of up to 70%.”

    Culpeper, VA — police department post

    Flock’s PIO kit / vendor FAQ language What Culpeper PD published
    “Flock Safety ALPR cameras help law enforcement investigate crime by providing objective evidence.” “The Flock Safety system also helps law enforcement solve crime by providing the objective evidence needed for investigations.”
    Vendor marketing claim repeated nationwide: “crime reductions of up to 70 percent.” “Communities using Flock Safety ALPR have reported crime reductions of up to 70%.”
    Vendor FAQ stock line also on Florence / Foothill–De Anza pages: “7 in 10 crimes are committed with a vehicle.” “According to the International Association of Chiefs of Police, 7 in 10 crimes are committed with a vehicle.”
  3. Flock buys labeled ads that look like articles in police trade media. The International Association of Chiefs of Police (IACP) — a big police professional group — has run “Sponsored Content” promoting license-plate readers and Flock’s talking points to chiefs and city decision-makers. That is paid advertising with a disclosure label, not independent journalism.

Why this matters in Crossville: if the main thing you hear is “it finds missing people,” ask where that story came from — a company-paid wire, a police press release built from Flock’s templates, or a sponsored trade ad. Then ask City Hall for Crossville’s own search rules, monthly use reports the public can check, and a yearly public yes-or-no vote before any renewal (no automatic renewal that skips a public vote).

Sources & archive links: Full source list for this section · GlobeNewswire — AMBER Alert partnership (2021) · GlobeNewswire — six abducted children (2026) · EFF — The Rise of the Police-Advertiser · Techdirt — companies writing press releases for cops · IACP — Sponsored Content (ALPR / Flock) · Campbell, CA — official Flock page (near word-for-word PIO lines) · Florence, OR — Flock install press release (PDF) · Foothill–De Anza PD — ALPR page · Culpeper, VA PD — Flock launch post

Open livestreams: no login, found via commercial search engines

In late December 2025, researchers and journalists documented a real security failure that undercuts “secure company cloud” talking points: Flock Condor people-tracking PTZ cameras were left streaming on the open internet with no username, no password, and no encryption. Anyone who found the URL could watch live video, download roughly 30 days of archive, and reach administrative/debug panels.

  • How they were found: Technologist Benn Jordan located admin interfaces via a commercial device search engine (documented as Shodan queries such as “Flock Admin”). Researcher Jon “GainSec” Gaines expanded the set with Shodan and ZoomEye to about 67 live feeds and debug interfaces — see his full write-up. Jordan’s report describes finding the interfaces “using a commercial search engine” — not hacking into a locked system.
  • Scale: 404 Media verified at least ~60 Condor cameras exposed nationwide; GainSec catalogued 67. Feeds covered playgrounds, parking lots, bike paths, street corners, retail lots, and more. The Verge summarized the same findings.
  • Independent proof: 404 Media’s Jason Koebler stood under a Condor at Harris Road & Young Street in Bakersfield, California, waved, and watched himself on the open livestream while colleagues watched remotely. Jordan filmed himself on Peachtree Creek Greenway feeds that auto-tracked people.
  • What was open: live stream + ~30-day archive download + settings/logs/diagnostics — “like Netflix,” in Jordan’s framing, with the ability to delete footage from the open admin UI (he reported not doing so).

“On my phone, I am watching myself in real time as the camera records and livestreams me—without any password or login—to the open internet.”

— Jason Koebler, 404 Media (Bakersfield Condor verification)

Flock called it a “limited misconfiguration” on a “very small number of devices” and said it was remedied; some agency emails blamed cellular/SIM testing configurations (WFLX / company confirmation). That does not erase the record: people-tracking cameras sold for public safety were briefly a public webcam network discoverable with commercial search tools.

Questions for renewal: Which Flock models are deployed or planned (Falcon vs Condor vs others)? Has the city received written notice of any exposure/incident affecting Crossville or Tennessee partners? What vendor security attestations and breach-notification terms are in the live contract?

Benn Jordan — Condor livestream exposure (YouTube) · 9NEWS short · 9NEWS video page

Sources for this section: Sources — Open livestream exposure (also listed under All cited sources).

51 cybersecurity findings (GainSec whitepaper)

In November 2025, independent researcher Jon “GainSec” Gaines published a formal whitepaper — Examining the Security Posture of an Anti-Crime Ecosystem — consolidating device-level research on Flock’s public-safety stack (license-plate readers, audio/gunshot hardware, compute modules, and related Android apps). The current public release documents 51 findings, including 22 assigned CVEs and additional findings pending CVE assignment, covering authentication, cryptography, and system-design issues (GitHub archive · GainSec summary).

This is separate from the Dec. 2025 open Condor livestream exposure. The whitepaper is about weaknesses in the product line itself — the kind of cameras cities lease — not a claim that Crossville’s units were exploited.

  • Hardcoded credentials: CVE-2025-47823 describes a hardcoded system password on Flock LPR firmware through version 2.2 (NVD / MITRE).
  • Unauthenticated admin API: CVE-2025-59403 is a publicly catalogued Flock camera admin-API issue also cited in this site’s liability table.
  • Device-level control (physical access context): GainSec’s series documents root/debug access paths on Falcon/Sparrow LPR units and related hardware when an attacker can reach the device. Flock’s public response stresses that exploitation would require physical access plus intimate knowledge of internal hardware, and says findings were disclosed earlier in 2025 with CVE registration and ongoing remediations ( Flock response).
  • Account / MFA risk (oversight, not a GainSec device CVE): A Nov. 2025 congressional letter asked the FTC to investigate Flock’s cybersecurity practices, including that MFA was not required and that weak MFA methods remained supported ( Wyden–Krishnamoorthi FTC letter, PDF).

Flock states the compiled findings do not impair customers’ public-safety use, that no customer action is required because Flock owns/maintains the devices, and that remediations continue via hardware and software updates. Residents can still ask whether Crossville received written notice, what firmware/security posture applies to local units, and what breach-notification terms sit in the live contract.

Questions for renewal: Which camera/compute models are installed? Has the city received written notice of the GainSec / CVE findings or related customer advisories? What firmware versions and MFA requirements apply to Crossville PD accounts? Who verifies evidence integrity if a device were physically compromised?

Primary sources: Sources — 51 cybersecurity findings · Secondary fact sheet: DeFlock Birmingham — The Facts.

CEO Garrett Langley: the trust problem

Garrett Langley, CEO of Flock Safety, speaking in a Forbes interview
Meet the CEO of Flock Garrett Langley Founder & CEO · Flock Safety
Still from the Forbes interview in which Langley called DeFlock a “terroristic organization.” Forbes interview on YouTube (used here for commentary on those remarks).
Flock Safety ALPR camera with solar panel mounted on a light pole
A Flock ALPR unit on a pole — the kind of hardware cities usually lease through a subscription, not buy outright. Flock typically keeps ownership of the camera. Photo: Wikimedia Commons (CC0).

Crossville’s contract is with a private vendor whose CEO has a public record residents can evaluate before renewal: an August 2025 company statement acknowledging that some public remarks about federal access were inaccurate and that Flock “communicated poorly,” followed by later quoted comments that cast FOIA requests as a “weapon” and called camera-mapping activists a “terroristic organization.” That is not “anti-police.” It is a vendor-trust issue before any renewal.

1. Acknowledged inaccurate public statements about federal access

After earlier public denials of federal contracts, Langley stated in writing that Flock had run limited pilots with U.S. Customs and Border Protection (CBP) and Homeland Security Investigations (HSI), and that some public statements had been inaccurate. Illinois’s Secretary of State said Flock’s handling of plate data violated state law. If the CEO’s public assurances can be “inadvertently… inaccurate” on federal access, Crossville should not take verbal promises on sharing or retention at face value — publish the contract, audit logs, and agency list.

“…some of our public statements inadvertently provided inaccurate information… we have engaged in limited pilots with the U.S. Customs and Border Protection (CBP) and Homeland Security Investigations (HSI)… We clearly communicated poorly.”

— Garrett Langley, Flock Safety statement (Aug. 2025)

National Lookup Tool: Flock told Congress that about 75% of its law-enforcement customers enrolled in the “National Lookup Tool,” which lets other enrolled agencies search opted-in camera data nationwide (agencies generally must share their own data to use the tool). Crossville has not published whether it is enrolled. Ask under ask before renewal.

Sources: NPR · ACLU — credibility / false claims · NBC Chicago (Illinois SOS) · National Lookup / cybersecurity oversight: Wyden letter to Flock (PDF) · Wyden–Krishnamoorthi letter to FTC (PDF) · Secondary fact sheet: DeFlock Birmingham — The Facts.

2. Told police chiefs that FOIA and resident questions are a “coordinated attack”

In December 2025 Langley emailed Flock’s law-enforcement customers that Flock and partner agencies were under “coordinated attack” from activists who want to “defund the police, weaken public safety, and normalize lawlessness” — and that they were “trying to turn a public records process into a weapon.” Staunton, Virginia’s police chief answered that residents seeking answers were “democracy in action.” The city then cancelled Flock because the CEO’s narrative “does not reflect the city’s values.”

“Let’s call this what it is: Flock, and the law enforcement agencies we partner with, are under coordinated attack… from the same activist groups who want to defund the police, weaken public safety, and normalize lawlessness… They’re also trying to turn a public records process into a weapon against you and against us.”

— Garrett Langley, Dec. 2025 email to police clients

“What we are seeing here is a group of local citizens who are raising concerns… In short, it is democracy in action.”

— Staunton Police Chief Jim Williams, reply to Langley

Sources: ACLU — CEO email · City of Staunton official cancellation · Augusta Free Press · NPR

3. Called camera-mapping privacy activists “terroristic”

In the same Forbes interview cycle, Langley described DeFlock — volunteers mapping publicly visible ALPR locations — as a “terroristic organization,” closer to Antifa than to lawful advocacy. Mapping where tax-funded cameras sit is not terrorism. It is the transparency Crossville is being asked to provide before renewal.

“…unfortunately there’s terroristic organizations like Deflock, whose primary motivation is chaos. They are closer to Antifa than they are anything else…”

— Garrett Langley, Flock Safety CEO (Forbes interview clip)

Watch Langley say it himself · full interview context: Forbes interview (timestamp)

After backlash he told Forbes (July 2026) the comments “were a mistake” and apologized. The video is still the record.

Sources: Forbes (Sept. 2025 profile) · Forbes (July 2026 apology) · Wikipedia — Flock Safety · Video clip

Why this matters in Crossville: a vendor that misstated who can access the network, then framed public-records requests and camera maps as extremism, is asking the council for trust. Residents should fill council meetings and demand cancellation — or at minimum a yearly public renewal vote with no automatic renewal that skips a public vote, monthly detailed use reports, a written policy, and a list of who can share the data.

This is not just about Flock

Canceling a Flock contract can help. It does not automatically end mass license-plate surveillance. What keeps happening: cities cancel Flock, then hire Axon (or shop Motorola Vigilant, Rekor, Genetec, Leonardo ELSAG, and others) to put plate readers on the same poles. The sales pitch changes. The job stays the same — scan every car, store the data, search it later. Some vendors go further: Leonardo pitches sensors that correlate plates with phone and Bluetooth IDs.

Documented July 2026: Gadget Review — “Cities Are Ditching Flock Safety Cameras – Then Hiring Axon to Do the Same Job” (Al Landes, July 20, 2026): cities drop Flock over data sharing, false alerts, and mass scanning, then hire Axon for the same kind of cameras. The company name changes. The surveillance does not go away.

Cities that canceled Flock, then hired Axon

Place What changed Source
Denver, CO Flock contract ended; City Council narrowly approved Axon — 50 cameras, $150k / 1 year. Officials said the new deal had tighter privacy rules Denver7 · Denverite
Douglas County, CO About $22.8M Axon deal: 100 Outpost plate readers (replacing 50 Flock cameras) plus countywide drones. Officials said “we own the data” Denver Gazette
Grand Chute, WI Board canceled Flock and signed Axon the same night; 23 cameras stay, swap by Sept. 1 FOX 11 / WLUK
Tempe, AZ 60-day free Axon Outpost trial after other Arizona cities canceled Flock ABC15
Syracuse, NY City Flock readers removed mid-July 2026; multi-year Axon plate-reader contract already approved — vendor swap, not an end to scanning National cases spotlight · Spectrum News 1
Knoxville, TN (nearby) After Flock ended, the city pushed a large Axon deal with a real-time crime center and drones — same kind of surveillance, different company. Axon also runs a vendor-styled “Connect Knoxville” private-camera program (see below) WVLT · Compass Knox · Connect Knoxville · Politician positions

Knoxville already went through this. Crossville should not look for a “better” plate-reader company. Take the cameras down. If police need location data on a specific car in a specific case, they should get a warrant — not scan every driver every day.

Axon and Flock were partners — not always rivals

Most people treat Axon and Flock as separate vendors. They compete for contracts now. For years they were tightly linked:

  • April 2020: Axon announced a commercial partnership and a minority, non-controlling investment in Flock Safety; Axon’s sales team would help sell Flock cameras to police (Axon investor release).
  • June 2021: Axon Fleet 3 in-car cameras launched with two-way Flock integration. Mobile ALPR reads from Fleet 3 appeared alongside fixed Flock cameras inside Flock’s Lookup Experience — one search covering both systems.
  • ~2023: The partnership dissolved after Axon launched its own competing ALPR line. Flock’s public blog confirmed that Flock processed Fleet 3 footage, applied Vehicle Fingerprint technology, and made it searchable inside Flock ((Flock blog).

Switching company names does not erase that shared-platform history. Crossville should ask in writing what Axon products (if any) the city or county already uses, whether any past or present sharing with Flock Lookup was enabled, what data was retained, and under what policy.

Nearby warning: “Connect Knoxville” and fususCORE

In 2026, connectknoxvilletn.org began promoting a program asking Knoxville residents and businesses to connect private security cameras to Knoxville Police. The civic-sounding .org site’s footer reads “Powered by Axon” / “©2026 Axon Enterprise, Inc.,” with contact at connect@fusus.com. The shop sells Axon fususCORE hardware that plugs into an owner’s camera system and can give police live access under owner-set settings, feeding FususONE / KPD’s Real Time Information Center.

The site presents itself as a community program while identifying Axon in its footer and offering Axon hardware for sale: residents pay for gear that can extend the police camera network onto private property. Registry data on the site is described as non-public. AI-enabled CORE devices are marketed with preloaded search/alert profiles (the FAQ says no facial recognition, but may recognize weapons, vehicles of interest, etc.).

Why Crossville should care: Connect Knoxville is the template. If a “Connect Crossville” or county lookalike appears, demand open-meeting approval, published retention/access rules, and enforceable standards before the first device is sold — not after. Deeper East Tennessee writeup: MaryvillePrivacy.org — Who is Axon? · EFF on private cameras in real-time crime centers: EFF (May 2023).

Sources: Connect Knoxville · Camera integration / fususCORE · Privacy FAQs · Shop

Axon + Ring: doorbell footage into Evidence.com

In 2025, Axon and Amazon’s Ring announced a partnership letting agencies request Ring doorbell footage — routed through Axon’s evidence system (Axon blog). Participation is framed as optional, but footage that leaves a home for a vendor cloud can be stored and shared under police/vendor rules. Amazon’s Ring canceled a separate planned partnership with Flock Safety in February 2026 after backlash; the Axon–Ring pathway remains relevant wherever agencies run Axon Evidence next door (including Knoxville’s stack).

Axon corporate record residents can evaluate (attributed reporting)

Public reporting and court documents — not Crossville findings — that inform how residents should weigh Axon’s “responsible innovation” pitch:

  • AI Ethics Board resignation (2022): Nine of twelve members resigned after Axon announced Taser-equipped drones for schools, bypassing a board that had voted against proceeding. Resigning members said they had “lost faith in Axon’s ability to be a responsible partner” (Policing Project statement). The Markup later reported that former board materials were removed from Axon’s site and that Axon acquired drone firm Dedrone.
  • FTC VieVu antitrust: The FTC challenged Axon’s acquisition of body-camera rival VieVu as anticompetitive; Axon litigated FTC process authority. Municipal plaintiffs later sued alleging monopoly pricing; a federal judge allowed damages claims to proceed (see Cohen Milstein case page).
  • “Excited delirium” litigation strategy: Physicians for Human Rights and Reuters investigative reporting describe how the contested diagnosis appeared in hundreds of deaths after Taser use and how Axon (then Taser International) funded research and expert testimony used in related litigation. Major medical groups do not recognize it as a valid cause of death. This site does not make medical findings; the reporting is relevant to vendor trust.

East Tennessee deep dive with full source list: MaryvillePrivacy.org — Who is Axon?

Axon is a platform, not one camera

Each product layer can feed the same cloud. Agencies that start with Tasers and body cams often add in-car video, Evidence.com, Fusus real-time ops, fixed ALPR (Outpost / Lightpost), drones, AI Draft One, Ring requests, and resident fususCORE devices. The privacy risk is aggregation: searchable location history plus private feeds plus AI analytics without strict written limits.

  • TASER / conducted energy weapons
  • Body-worn cameras → Evidence.com
  • Fleet 3 in-car video + mobile ALPR (historically searchable inside Flock Lookup)
  • Fusus / RTCC real-time operating picture
  • Outpost / Lightpost fixed ALPR (sold as Flock replacement)
  • Axon Air / Dedrone aerial and counter-drone tools
  • Draft One AI report writing
  • Ring request pipeline + fususCORE private-camera hardware

Problems with Axon too

Some cities hire Axon after dumping Flock because Axon says it is safer or more private. That sales pitch does not stop abuse. Axon products raise separate transparency, conflict-of-interest, and false-alert concerns documented below. None of this is a reason to bring Axon — or any plate-reader company — to Crossville.

Axon · Draft One · AI police reports · 2024–2025

Draft One: AI writes the police report, then deletes the draft

Axon’s Draft One uses body-cam audio to write a draft police report. EFF and Ars Technica reported that the original AI draft is not kept, so it is hard to compare versions or prove what the AI wrote. Agencies can also turn off a check meant to make sure officers actually read the draft — also covered by Mother Jones and the ACLU’s Automated Police Reports brief (PDF). Generative-AI tools can produce inaccurate content. If the draft is gone, courts and defense lawyers have a harder time seeing what the machine said versus what the officer signed.

Sources: EFF — argues Draft One frustrates transparency · Ars Technica · Mother Jones · ACLU — Automated Police Reports

Axon Fusus · Atlanta ethics · conflict of interest

Atlanta ethics finding: police official owned a Fusus stake while working city business without required disclosure

The Atlanta Ethics Office found APD Chief Administrative Officer Marshall Freeman consulted for Fusus, owned a stake in the company, and worked on Fusus city business without proper disclosure — including pitching it to other agencies (Atlanta Press Collective; The Guardian). Axon later bought Fusus for about $240 million. The Intercept reported he pitched at least 14 cities; at least nine later signed or expanded Axon Fusus contracts. The Ethics Office fined him $5,000.

Sources: The Intercept — “Duped” · The Guardian · Atlanta Press Collective · Atlanta Ethics Office decision (PDF)

Axon Fleet 3 plate reader · false alerts

Axon’s help pages describe false alerts and outdated stolen-car lists

The Fleet 3 ALPR FAQ tells officers how to mark alerts as Incorrect or Wrong State/Province. Axon’s snooze guidance describes procedures for handling incorrect and wrong-state alerts: older software often cannot tell which state a plate is from (so the same characters from another state can trigger a false alert), can sometimes “read” things that are not plates, and depends on stolen-car lists that can be out of date. Axon’s early 2026 firmware notes describe improved state detection aimed at reducing Wrong State hits. False or outdated alerts are how innocent people get stopped at gunpoint when officers do not check the plate themselves.

Sources: Axon Fleet 3 ALPR FAQ · Axon — Snooze / false-alert guidance · Axon Jan 2026 release notes — state recognition

Multiple plate-reader companies · LAPD Inspector General · 2026

LAPD: about 1 in 3 stolen-car alerts were wrong. Axon cameras were part of the system

LAPD’s Inspector General looked at two months of alerts. Officers treated 161 alerts as real stolen-vehicle hits that turned out wrong — often because the stolen-car list was outdated — then ran high-risk stops with guns drawn (TechTimes; Biometric Update). Reporting on the audit says LAPD’s plate-reader system included Flock, Motorola, and Axon. LAPD later let the Flock contract expire. Changing companies does not fix bad stolen-car lists, skipped checks, or scanning every car. Taking the cameras down does.

Sources: TechTimes / OIG coverage · Biometric Update

Axon can be abused the same way Flock is

Documented Flock and ALPR problems — personal lookups with no case, misread plates, outdated stolen-car lists — are not unique to one company. They happen when police get a searchable log of where cars go, alerts that feel like “stop that car now,” and little day-to-day supervision. Axon’s in-car readers, pole cameras, Evidence storage, and Fusus real-time tools can do the same things — and can pull in body cams, private-camera feeds, and AI report writing too. That is why Crossville should not switch companies, and why Crossville should take these cameras down completely and stay away from the same kind of technology under any other name.

Same problems with Flock or Axon

  • Personal stalking: An officer looks up an ex’s plate or travel history — the Institute for Justice has documented dozens of these cases; which company sold the cameras does not matter much
  • Searches with no real case: Looking up yourself, family, or coworkers for no official reason — AJC / GBI reporting (2026) describes Georgia agencies finding misuse when they actually checked the logs
  • False or outdated stolen-car alerts: A misread plate or a car no longer stolen still triggers a stop if nobody checks
  • Wrong-state or similar-looking plates: Axon’s own docs describe Wrong State hits; Flock misreads are widely reported — both put innocent drivers at risk
  • Sharing with outsiders: Local data can still leave the agency through partners and APIs unless the cameras are gone
  • Logs that do not stop abuse: Search logs exist; misuse is usually found by victims or rare audits — not by company marketing

Axon adds more problems, not more privacy

  • More cameras, more places: Readers in patrol cars plus pole/streetlight cameras mean more of your trips get logged
  • Fusus real-time screen: Combines live video and plate hits; warrant rules still depend on agency policy and law
  • Evidence.com: Plate records sit with body-cam video in the same Axon system — harder for a city to leave Axon entirely
  • Draft One: AI writes report drafts, then deletes them — harder to see what the machine produced
  • Ring + fususCORE: Doorbell requests and resident-purchased private-camera hardware extend the network into homes (Ring, Connect Knoxville)
  • Bigger packages: Places like Douglas County swap Flock for Axon and add drones — more surveillance while saying “we canceled Flock”

Other vendors shopping the same job

Axon is the loudest Flock replacement in 2026 — not the only one. Procurement writeups and vendor pitches show agencies evaluating Motorola Vigilant, Rekor, Genetec AutoVu, Verkada, Avigilon, PlateSmart, and Leonardo ELSAG alongside or instead of Flock. Different logos. Same mass plate capture unless Crossville refuses the category.

Leonardo ELSAG · SignalTrace · June 2026 reporting

Leonardo SignalTrace — plate readers that also scoop phones and Bluetooth IDs

Leonardo (ELSAG ALPR line) markets SignalTrace as a retrofit for license-plate cameras: sensors capture identifiers from nearby phones, wearables, Wi-Fi sources, RFID tags (even pet microchips in product materials), and vehicle systems, then correlate recurring “electronic fingerprints” with plate reads and locations. Company materials say the system can help identify a vehicle even without a readable plate. Reporting by 404 Media and Techdirt describes this as turning car trackers into people trackers — mass collection first, investigative query later.

Why Crossville should care: a “not Flock” camera deal can still expand into device-sensing ALPR. Ask in writing whether any vendor has pitched SignalTrace, ELSAG, or Bluetooth/Wi-Fi vehicle correlation — and refuse the whole stack.

Sources: 404 Media · Techdirt · Leonardo — ELSAG SignalTrace product page · The Deep Dive

Motorola Solutions · Vigilant / LEARN · long-running ALPR network

Motorola Vigilant — enterprise plate databases and sharing fights

Motorola Solutions acquired Vigilant Solutions and operates large ALPR networks and the LEARN-style sharing platforms used by many agencies. EFF’s ALPR research found agencies sharing Vigilant data with large numbers of other entities on average. In 2024, CISA published advisories on Vigilant ALPR security flaws (weak credentials / encryption issues). In July 2026, a California class action (amended complaint) alleged Motorola/Vigilant facilitated unlawful out-of-state and federal sharing of California ALPR data — allegations in a filed complaint, not a court judgment.

LAPD’s multi-vendor ALPR stack (Flock, Motorola, and Axon named in Inspector General coverage) shows cities already mix these brands. Quitting Flock for Motorola is still mass scanning.

Sources: EFF — Vigilant / LEARN sharing research · EFF — CISA / Vigilant vulnerabilities (2024) · Gibbs Mura — Motorola ALPR privacy lawsuit summary (July 2026 amended complaint) · Biometric Update — LAPD multi-vendor ALPR

Who else is in the procurement mix

Industry and procurement trackers list additional ALPR competitors cities evaluate when Flock is politically toxic. Treat every name as the same ask category unless Crossville refuses plate scanning entirely:

  • Axon — Outpost / Lightpost fixed ALPR, Fleet 3 mobile ALPR, Fusus RTCC, Evidence.com, Draft One, Ring / fususCORE (covered above).
  • Leonardo ELSAG — fixed/mobile ALPR plus SignalTrace device-correlation pitch (above).
  • Motorola Vigilant — enterprise ALPR + shared databases (above).
  • Rekor — often pitched for highway / transit / open-road ALPR.
  • Genetec AutoVu — ALPR inside Genetec video / security platforms used by larger cities.
  • Verkada / Avigilon / PlateSmart — cloud or software ALPR layers that can ride on existing cameras.

Secondary market overview (vendor-facing, use cautiously): Civic IQ — Flock alternatives list (2026). Crossville’s ask stays the same: no mass ALPR under any brand.

What to tell City Council

  • Cancel Flock and remove every camera. Do not evaluate, demo, trial, or install Axon, Leonardo ELSAG / SignalTrace, Motorola Vigilant, Rekor, Genetec, Verkada, or any other plate-reader system — ever again.
  • Do not scan every driver. Ordinary drivers’ plates should not be logged all day with no suspicion.
  • Warrants for specific lookups. If police need where a specific car went in a specific case, they should get a warrant — not a subscription that scans everyone first.
  • Claims like “the city owns the data” or “shorter retention” do not fix this. Taking the cameras down does.

News covering these company swaps: Gadget Review (July 20, 2026) · Nearby Axon package: Compass Knox — Knoxville Axon / RTIC

Email leaders · Misuse cases · Cancellations · Other vendors

In the media

Independent and local news coverage of Flock Safety ALPR cameras — how the product is marketed versus how it is used, wrongful stops from camera errors, security exposures, and police misuse. These clips are third-party reporting; they are not Crossville-produced content. Click play to load the YouTube player (YouTube is not contacted until you play).