In mid-August 2026, Flock Safety CEO Garrett Langley appeared on Fox News and told the country that when people focus on privacy or safety alone, they are ‘prioritizing the wrong thing.’ Within 24 hours, the clip went viral on X. Within days, users began posting what they claimed was his home address in Atlanta, some pairing it with blurred images lifted from Google Street View. The observation that followed: his call for compromise did not appear to extend to himself.
In this article
What Langley Said, and What Happened Next
On Fox News’ Saturday in America, Langley stated that “when people talk about just one of these, privacy or safety, they’re prioritizing the wrong thing, and what we have to prioritize as a country is compromise.” He was asked whether Flock’s cameras risked ushering in a surveillance state, and he rejected the framing outright.
The clip traveled fast, gathering thousands of critical replies on X within a day, where many users made the same point: that Flock’s cameras log the movements of millions of drivers who never agreed to be tracked. Some users began posting what they said was Langley’s home address, a few pairing it with blurred images lifted from Google Street View. One X user noted that the “compromise apparently doesn’t apply to him.”
The authenticity of the address could not be confirmed, and Newsweek declined to publish or link to it. X’s moderation team began removing posts under its doxxing policy, but users continued re-uploading addresses with altered wording and images to evade automated filters.
Flock Safety, founded in Atlanta in 2017, now runs automated number-plate readers for more than 4,800 police forces across 49 states. As of 2025, Flock says it operates in over 5,000 communities and performs over 20 billion scans of vehicles in the US every month. A funding round in March 2025, led by Andreessen Horowitz, raised $275 million and valued Flock at $7.5 billion.
The Abuse Cases That Preceded the Fox Interview
Langley’s television appearance did not occur in a vacuum. In early August 2026, a Washington Post investigation documented dozens of cases in which officers had used the Flock network to track partners, former partners, or relatives. One example: a former Milwaukee police officer whose court documents showed he secretly used Flock to search the location of his romantic partner 124 times.
A separate tally from Washington Post reporting identified 46 cases of police misusing Flock’s surveillance technology. Other reports put the number near 50. Articles about law enforcement officers abusing ALPR surveillance footage to spy on their wife or scorned lover have become a “regular occurrence,” according to reporting from Deseret News.
On August 13, 2026 – days before Langley’s Fox appearance – Flock announced it would implement several new safeguards, including shrinking the default data retention period from one month to one week. Both of these changes can be overridden: police can save data for a longer time period by using a setting called Evidence Mode. Existing customers retain their current retention periods unless they choose to opt into the new seven-day standard.
The American Civil Liberties Union called the changes a “step in the right direction” but found that overall they were “merely retreads of previous inadequate safety measures.” Jay Marlow of the ACLU noted that while he welcomed the shortened retention time, a week is still short of the ACLU’s recommended 48 hours, and the company’s reforms did little to address his broader concerns with license plate tracking.
How a Local Tool Became a National Database
Flock’s model depends on aggregation. Law enforcement offices that contract with Flock do not need a warrant to access travel information and can opt to share their data with local law enforcement offices, statewide or nationwide, creating a database of Americans’ real-time movements accessible to a range of police officers that spans much of the country.
Customs and Border Protection reportedly had access to tens of thousands of Flock units, a scale that has digital rights advocates describing the network less as a neighborhood safety tool and more as a national vehicle-movement database. This is how a local camera network becomes a de facto national surveillance system without Congress ever voting on it. The data sharing is voluntary and legally murky. There is no federal statute authorizing it.
DeFlock.org, which crowdsources Flock camera locations, has mapped over 124,000 suspected LPR deployments across the United States. The coverage in urban and suburban areas is dense enough that driving across most American cities generates a near-continuous surveillance record.
The scale changes the character of the tool. What Flock markets to municipalities as a targeted investigative aid becomes, in aggregate, the infrastructure for tracking vehicle movements at a population scale. Senator Ron Wyden of Oregon wrote in October 2025 that “abuse of Flock cameras is inevitable, and Flock has made it clear it takes no responsibility to prevent or detect that.” He recommended that communities that have installed Flock cameras “reevaluate that decision.”
Bipartisan Resistance and Municipal Cancellations
The backlash to Flock is not confined to one political faction. Three House Republicans recently introduced a bill that would prohibit the federal government from purchasing automated surveillance systems that use facial recognition, biometric IDs, or license plate recognition, “including a Flock Safety camera.” Representative Keith Self quoted Benjamin Franklin and introduced the PRIVACY Act, which would require federal agencies to obtain warrants before querying ALPR data. His statement: “The Fourth Amendment is not up for ‘compromise.'” Senator Josh Hawley opened a formal congressional investigation into Flock’s data practices.
More than 50 cities and towns, spread across some 20 states, have moved to cancel or switch off their Flock contracts in 2026, according to activist trackers. Santa Cruz’s City Council voted in January 2026 to end their contract with Flock Safety. Mountain View’s Chief of Police announced in February 2026 that they would immediately discontinue use of Flock ALPR cameras. Chandler City Council announced in August 2026 that it would stop using Flock cameras and would no longer pursue a renewal of their contract.
Utah Governor Spencer Cox said he was “deeply concerned” about recent news stories where law enforcement abused the surveillance technology, but added that Utah has “been in the forefront of really trying to put guardrails around the use of those types of technologies to protect citizens’ privacy.” Cox directed the state’s Privacy Commission to review the rules governing the cameras after a Utah Civic Compact investigation found that 10 cameras in a single Utah county were searched 5.1 million times by police agencies across 45 states.
Architectural Lessons for Email and Communication Infrastructure
The trajectory of Flock Safety offers a structural parallel to email. Both are ostensibly decentralized systems – email servers operated by independent entities, cameras deployed by individual municipalities – that aggregate into centralized surveillance infrastructure when data-sharing becomes the default posture.
The privacy failure is not in the camera itself but in the aggregation architecture. A municipality installs cameras to track stolen vehicles. Those cameras feed a regional network. That network shares with federal agencies. At no point does a single decision create a national vehicle-movement database, yet one emerges incrementally, each step justified by the step before.
Email operates under similar pressure. A service begins by offering encrypted storage. Then it adds spam filtering, which requires scanning message content. Then it introduces AI features, which require broader content access. Then it partners with advertisers or offers enterprise tools that require new retention policies. At each stage, the privacy model degrades slightly, and at no stage does the user make a clear decision to participate in the final architecture.
The defense against this pattern is not voluntary safeguards – Flock’s policy changes are reversible by design – but architectural constraints that prevent data aggregation in the first place. For email, this means end-to-end encryption that the provider cannot bypass, local storage that the provider cannot access, and transport protocols that do not require centralized routing. Tor-routed email with client-side encryption and no server-side metadata retention represents one such architecture. Messages encrypted on the client, routed over Tor, leave no metadata trail after delivery because the server constructs no movement log.
Post-quantum cryptography extends this principle. Open-source lattice-based encryption – such as PQCServer, released under AGPL-3.0 – implements algorithms resistant to quantum decryption. This matters not for today’s threat model but for the retention model: if data encrypted today can be decrypted in 2035 when sufficiently powerful quantum computers exist, then retention itself becomes a vulnerability. Encryption that remains secure against future capabilities allows for architectural designs where old data poses no risk even if retained.
What the Street View Blur Actually Reveals
The detail that circulated most widely – that Langley’s home appeared blurred on Google Street View – functions as a meme, not an argument. Google allows any property owner to request blurring. Thousands of people do this. Langley’s decision to blur his home does not prove hypocrisy; it proves that he values the privacy afforded by obscuring his residence from public view.
The observation is nevertheless structurally significant. Langley built a business on the premise that capturing and indexing vehicle movements in public spaces is an acceptable tradeoff for public safety. The fact that he exercised available privacy protections for his own home – protections unavailable to the millions of drivers whose movements Flock logs – does not invalidate his business model, but it does clarify it. Privacy is a resource that can be selectively allocated. Those who control surveillance infrastructure allocate it to themselves. Those who do not, cannot.
This is the compromise Langley described: a system where privacy is negotiable, subject to municipal contracts, law enforcement discretion, and vendor policy updates. The alternative architecture – one where privacy is a structural property of the system rather than a policy parameter – does not require compromise because it does not create the tradeoff. End-to-end encrypted communication over Tor does not generate the data that would need to be protected by policy. It does not generate it in the first place.
Flock’s municipal cancellations, congressional inquiries, and contractual rollbacks will continue. The company will adjust retention periods and add audit features. Municipalities will debate the correct balance. None of this addresses the architectural issue: a system designed to aggregate surveillance data will aggregate surveillance data. Voluntary safeguards do not change the design. They govern its operation.