Flock Safety lays off 18% of its workforce as backlash grows against its AI surveillance cameras

By Zendoric · October 10, 2026.

Flock Safety, the Atlanta company that operates one of the largest license plate reader camera networks in the United States, is preparing to lay off some 270 employees —18% of its workforce of roughly 1,500 people— according to sources with direct knowledge of the plan cited by Reuters in an exclusive. Those affected will leave the company at the end of October.

The company prefers a different name for the move. Its spokesperson, Paris Lewbel, told TechCrunch it was a "generous Voluntary Separation Program" and not a layoff, and declined to say whether CEO Garrett Langley or other executives took pay cuts as part of the plan.

The cut comes amid public scrutiny of its business. Flock's network comprises some 120,000 AI-equipped cameras installed on streets and highways across 49 states, automatically reading license plates and logging vehicles as they pass; the company sells that system and the associated software to more than 4,800 police forces and close to 1,000 businesses. Florida banned the use of these readers on its state highways in September on privacy grounds, and Flock faces a lawsuit in Virginia arguing that the mass deployment of its cameras amounts to warrantless surveillance, according to Reuters. Home Depot investors have also called for a review of the retailer's agreements with surveillance technology providers after it emerged that Flock data was used in immigration enforcement. TechCrunch has also reported cybersecurity failures that left cameras on the network exposed on the internet, a finding none of the other outlets compared picked up.

Its financial backing contrasts with that climate. Founded in 2017, Flock has raised more than $950 million in venture capital, including a $275 million round led by Andreessen Horowitz in March that valued the company at $7.5 billion. A Reuters/Ipsos poll puts public support for its cameras at 38%, against 47% opposed, and Reuters already describes the issue as a political weapon heading into November's midterm elections.

What sets this case apart is who the cut hits. Most of the AI-and-jobs news we cover involves companies using AI to shed their own staff. Here the opposite is happening: an AI company is cutting its own workforce because the product it sells —persistent perimeter surveillance wired into police and, according to the complaints, into immigration enforcement— has run into a social limit that neither capital nor growth could buy off. The money bought scale; it did not buy social license for that scale.

That limit, moreover, was not set by an AI law. It was set by a state legislating over its own roads, a lawsuit invoking constitutional guarantees that already existed, and shareholders pressuring a corporate client. It is the same pattern we have been observing on other AI regulatory fronts in the United States: the effective brake comes from the law and the institutions that were already there, not from rules tailor-made for algorithms. For a technology that combines automatic recognition, police databases and national scale, that is probably the fastest control mechanism available, while the specific legislative debate on AI lags behind.

Our reading is that this episode does not contradict the underlying thesis that AI, properly directed, can free up time and resources for people: what it illustrates is that the promise depends on what gets deployed and with whose consent. An AI that speeds up the diagnosis of a disease or makes energy cheaper builds social trust in the whole technology; an AI that normalizes the permanent tracking of vehicles without citizens getting a say erodes it, and that erosion ends up taking its toll —in state bans, lawsuits and, now, layoffs— even on AI applications that do deserve the benefit of the doubt. Flock is not announcing the end of algorithmic surveillance: with 4,800 police customers, its product will keep growing in many places. But it does confirm that, in the short term, the main risk for AI companies is not only technical or strictly regulatory: it is one of legitimacy, and it is measured in polls, in courts and, as of this month, in payrolls.