The Facts

About Flock Safety

Flock Safety is the brand of Flock Group Inc., a surveillance company founded in 2017 that sells automated license plate reader (ALPR) cameras and audio-detection devices to police departments, businesses, and HOAs, connected through a national searchable network. A major early investor, Founders Fund, is also an early Palantir investor. Fortune (Aug 24, 2026) reported Flock valued at about $8.3 billion, with more than $500 million in annual recurring revenue, deployed in 5,000+ communities across 49 states, running 120,000+ cameras, and processing roughly 20 billion reads per month.

Flock's own sales materials describe machine vision that identifies vehicles by make, model, color, and features such as roof racks, bumper stickers, and dents, and can capture a car "with the absence of a license plate." Flock trademarked this capability as Vehicle Fingerprint®. Its own LPR policy states that it keeps images "for the purpose of improving Flock Services through machine learning." The plate is one field in an AI-built profile of every passing car, not the whole product.

Murphy's Flock program: what's known, what isn't

What's confirmed: Murphy Police Department operates Flock ALPR cameras. The council first approved them on December 3, 2024, voting 5-0 to accept a Texas "Catalytic Converter Theft" grant (Senate Bill 224) to purchase Flock Safety cameras (Resolution 24-R-1115); Chief Jeff Gibson and a detective presented the item alongside Flock staff (City of Murphy Regular City Council Meeting minutes, December 3, 2024). On May 5, 2026, the council again voted unanimously to seek another SB 224 grant for fiscal year 2027, a Motor Vehicle Crime Prevention Authority program that funds catalytic-converter and auto-theft prevention and requires a local cash match, to keep the program funded (Murphy Monitor).

What has not been disclosed: the city has not published the Flock contract, the number of cameras, the locations, the annual cost, or the renewal terms. There is no camera count on the city's website, and Murphy does not appear on national trackers that document Flock deployments. Both approvals were grant votes rather than a regular budget line, and residents themselves were never asked to vote on whether Murphy should run license-plate surveillance.

Where this page says "undisclosed" or "not public," that means exactly that: no verified figure exists in the public record as of this writing. We do not estimate or guess at camera counts or costs.

Documented security failures

In December 2025, technologist Benn Jordan and 404 Media found at least 60 Flock cameras exposed on the open internet with no password, streaming live video and 30 days of archived footage, including footage of a playground. Researchers have separately found Flock and law-enforcement login credentials for sale on dark-web markets.

A 2025 Illinois Secretary of State audit found that Flock gave U.S. Customs and Border Protection access to Illinois data in violation of state law, through a pilot program Flock's own leadership said it was unaware of. The pattern across these incidents: every safeguard Flock has added arrived only after it was caught, not before.

Effectiveness and accuracy

Flock advertises 96% character-level read accuracy "under optimal conditions." Independent audits tell a different story. Roseville, California found 71% of its own Flock alerts were wrong over a two-year period. The LAPD Inspector General found 161 of 498 hot-list alerts, roughly one in three, were false, and LAPD let its Flock contract expire the day after the audit came out. A Vallejo, California study found up to 37% of fixed-reader hits were erroneous. When the ACLU reviewed nearly 300 police departments, some reported hit rates as low as 0.01% (Rhinebeck, NY) and 0.08% (High Point, NC).

By Flock's own numbers, across roughly 20 billion monthly reads nationally, even a small error rate implies hundreds of millions of misreads every month. Misreads have real consequences: Amber Newell, a Wisconsin mother, was held at gunpoint after a misread; an Aurora, Colorado family, including four children, was handcuffed at gunpoint over a plate that actually belonged to a Montana motorcycle, a case that settled for $1.9 million. Flock's own effectiveness studies are commissioned by Flock; independent research on outcomes is thin and mixed.

The Fourth Amendment and the law

A single public photograph of a car is settled law. But ALPR networks assemble a searchable, weeks-long history of everywhere a vehicle has been, queryable by thousands of agencies. In Carpenter v. United States (2018), the U.S. Supreme Court held that an aggregated record of a person's movements requires a warrant when it comes from a cell phone. The Institute for Justice's federal lawsuit over Norfolk, Virginia's Flock network, Schmidt v. City of Norfolk, is on appeal at the Fourth Circuit. No court has finally ruled ALPR networks unconstitutional, and Texas has no state law limiting ALPR use.

For comparison: Texas banned red light cameras in 2019 (HB 1631), largely over due-process concerns, because those cameras issued fines, which gave drivers standing to sue. Flock cameras never ticket anyone; they log every passing car quietly, so almost nobody has had standing to challenge them in court. "Legal" here means untested, not vetted.

Where the data goes

Flock's terms grant Flock an "irrevocable, perpetual, worldwide" license to use customer data to "support and improve Flock's products and services," a license that survives even after a city cancels its contract. The Texas Tribune (Aug 24, 2026) reported that Flock retains this perpetual right to use customer data with no expiration after termination. Fortune (Aug 24, 2026), citing an LAPD Inspector General audit, reported that Flock's contracts let it use collected data "for any purpose in Flock's sole discretion," including to train machine-learning algorithms.

Because Flock operates a national network, data captured by any one city's cameras is searchable by other agencies participating in the network, and in documented cases, by federal authorities. That is a designed feature of the product, not an accident.

National context: misuse, cancellations, bipartisan alarm

At least 27 officers across 18 departments in 9 states have been fired, charged, or disciplined for abusing ALPR systems, most often to track women they knew; roughly ten more are under investigation, and national tallies of alleged cases run from 50 to more than 70. In August 2026, an Angelina County, Texas grand jury indicted a Lufkin police officer on 100 felony counts for using Flock to track eleven people; the department shut its own cameras off in response. Four Savannah, Georgia police employees were arrested the same week.

As of August 21, 2026, DeFlock's national tracker counted 106 communities that had canceled, rejected, deactivated, paused, or banned Flock or ALPR programs (46 canceled, 20 rejected, 19 deactivated, 8 removed, 12 paused, 1 ban), with at least a dozen more documented exits not yet logged, putting the true number above 115. Communities in 23 states acted in 2026 alone. In Texas that includes Hood County, Guadalupe County, Bandera, Lago Vista, San Marcos, Austin, and, on August 25, 2026, neighboring Wylie, whose council voted 7-0 to end its Flock contract after a resident campaign; Wylie's cameras come down when its contract expires October 13, 2026.

This is not a partisan issue. In August 2026, Senators Bernie Sanders (I-VT) and Rand Paul (R-KY) both condemned Flock's network as mass surveillance within days of each other. Two federal bills restricting ALPRs are Republican-sponsored: the PRIVACY Act (Rep. Keith Self, R-TX) and H.R. 9800, the Protection Against Mass Surveillance Act (Rep. Tim Burchett, R-TN). A congressional investigation into Flock is Democratic-led. Flock's own CEO, Garrett Langley, told Fox News and TechCrunch (Aug 23, 2026) that "there's no regulation, there's no accountability, and we think that's wrong," while calling for "compromise."