The pitch was simple: give police departments a network of automated license-plate readers, let the cameras do the watching, and crime goes down. Flock Safety, the Atlanta-based surveillance company, built its business on that promise — and on the implicit assurance that the technology would be used as advertised. By 2024, Flock had placed roughly 80,000 cameras across the United States, making it one of the largest private contributors to American law enforcement infrastructure in the country's history.
What the company did not build, according to an investigation by The Lever, were meaningful safeguards against the officers who control those cameras. Dozens of police officers have since been accused of improperly accessing the system — not to solve crimes, but to track ex-partners, surveil civilians with personal grievances against them, and conduct unauthorized searches of people's movements. The cameras watched. Nobody watched the watchers.
This is not a story about a few bad actors who broke the rules. It is a story about a surveillance architecture that was designed, from the beginning, to trust police departments with nearly unchecked access to one of the most intimate data streams imaginable: where your car is, when, and how often. The abuse was not a bug in Flock's system. It was a foreseeable consequence of how the system was built.
License-plate readers have been a fixture of American policing for decades, but Flock Safety industrialized the concept. The company, founded in 2017, grew explosively by offering departments a subscription-based model: cameras installed at low upfront cost, data stored in the cloud, and search tools that let officers query the movement history of any vehicle that passed a Flock camera. The system is interoperable — meaning an officer in one jurisdiction can, depending on data-sharing agreements, access reads from cameras in another. A single officer with access credentials holds a window into an enormous amount of daily American life.
The appeal to police departments was genuine. Flock cameras have contributed to solving vehicle thefts, locating missing persons, and identifying getaway cars. The company's marketing emphasized these wins relentlessly. What the company emphasized far less — what it apparently did not build robust infrastructure around — was the question of what happens when an officer queries the system for reasons that have nothing to do with crime. When the ex-girlfriend drives by the station. When the neighbor who filed a noise complaint keeps showing up in the feed. When the system becomes a personal tool rather than a public safety one.
Surveillance technology and intimate partner abuse have a well-documented relationship. The same tools that allow an officer to locate a stolen vehicle also allow that officer to track a former partner's daily movements — when she leaves for work, which route she takes, whether she's been to a new address. For survivors of domestic violence who have left relationships with law enforcement officers, the asymmetry is stark: they cannot know whether they are being watched, and by whom, and why. Flock's network, spread across tens of thousands of cameras in thousands of communities, extends that asymmetry nationwide.
The structural problem is one of access architecture. Most institutions that handle sensitive data — medical records, financial information, legal communications — operate on a need-to-know principle enforced by logging, auditing, and oversight. An employee who pulls a patient's records without clinical justification leaves a trail that can be investigated. The same principle applies, in theory, to law enforcement databases. In practice, audit cultures inside police departments are notoriously weak, and Flock's system depends almost entirely on departments to police themselves. The company can see that a query was made. Whether that query was legitimate is a question it largely leaves to the agency that made it.
Flock Safety sells subscription-based license-plate reader cameras to police departments and private communities. Each camera captures images of passing vehicles, extracts plate numbers, and stores the data in a searchable cloud database. Officers with credentials can query movement histories for any vehicle captured in the network. Data-sharing agreements between jurisdictions can dramatically expand the geographic range of any single officer's search capability.
This is not a hypothetical risk. The Lever's reporting documents dozens of officers accused of improperly accessing the system. The cases span multiple states and departments. The pattern is consistent: an officer with a personal grievance, a terminated relationship, or a vendetta uses their professional credentials to conduct surveillance that has no law enforcement justification. The cameras, designed to watch streets, end up watching specific people — people who have no way of knowing it is happening.
For communities that have spent years pushing back against the expansion of surveillance infrastructure, the Flock cases confirm a concern that critics raised when these cameras first started appearing. Surveillance tools sold to police departments on the basis of public safety benefits do not stay confined to their advertised use cases. They expand. They get repurposed. They get abused by the people who control them, because the people who control them have broad discretion and limited oversight. The technology does not create the impulse to abuse power — but it dramatically expands what that abuse can look like.
The domestic violence dimension deserves particular attention. Research consistently shows that intimate partner violence is disproportionately prevalent within law enforcement communities, and that victims in those relationships face compounded barriers to safety: fear of retaliation, skepticism from other officers, and the knowledge that their abuser has professional tools and professional networks that ordinary abusers do not. Flock's network adds another dimension to that asymmetry. An officer who wants to track a former partner's movements no longer needs to physically follow her. He can query a database. As Tinsel News has reported previously, the fusion of surveillance technology and institutional access has become one of the defining civil liberties problems of the current era — and the Flock cases are a version of it playing out inside police departments across the country.
The accountability question is complicated by the nature of Flock's business model. The company sells to agencies, not to individuals. When an officer misuses the system, the formal liability sits with the officer and, potentially, the department — not with Flock. The company's terms of service prohibit unauthorized access. Whether it has the technical means or institutional will to detect and flag that access systematically is a different question. Building genuine audit infrastructure into the system — automatic alerts when an officer queries a plate associated with a personal contact, mandatory justification fields for every search, third-party auditing of access logs — would cost money and might make the product less convenient. It might also make it harder to sell.
There is a broader market logic at work here. Flock Safety is not a public agency. It is a venture-backed company competing for law enforcement contracts in a market where ease of use and interoperability are selling points. Friction — the kind that comes from robust oversight requirements — is a competitive disadvantage. The company's incentives point toward making the system as accessible and searchable as possible for the officers who use it. The incentives do not naturally point toward building the safeguards that would protect the people being watched from the people doing the watching.
This is where the public interest and the private market diverge most sharply. Surveillance infrastructure sold to government agencies carries public safety implications that extend well beyond the terms of a commercial contract. When a police department installs Flock cameras, it is making a decision about what kind of power its officers will have over the communities they serve. That decision deserves public deliberation, oversight mechanisms with actual teeth, and accountability structures that do not depend on departments investigating themselves.
Some cities and counties have begun asking harder questions about the surveillance tools they purchase. Civil liberties organizations have pushed for surveillance technology ordinances that require public hearings before new systems are deployed, mandate regular audits, and establish clear penalties for misuse. California, Washington, and a handful of other states have passed legislation requiring local governments to approve surveillance technology before law enforcement can use it. These frameworks are imperfect and inconsistently enforced — but they represent the architecture of accountability that the Flock cases show is missing at the federal level.
The federal picture is bleaker. There is no comprehensive national framework governing how law enforcement agencies can use private surveillance networks, no federal requirement that access to systems like Flock's be logged and audited by an independent body, and no agency with clear jurisdiction over the intersection of commercial surveillance technology and police power. As Tinsel News has covered, the legal doctrine of qualified immunity already makes it extraordinarily difficult to hold individual officers accountable for civil rights violations — and the absence of federal surveillance standards means that the officers who abuse systems like Flock's are operating in a regulatory vacuum that compounds that protection.
The victims of these abuses — former partners tracked without their knowledge, civilians surveilled for reasons that have nothing to do with public safety — have limited recourse. They may not know they have been watched. If they do find out, the legal path to accountability runs through institutions that have structural incentives to protect their own. The cameras generate data. The data generates power. The power, in the absence of oversight, generates abuse.
What Flock Safety built is not, in itself, the problem. License-plate readers exist. The technology will not disappear. The question is whether a network of 80,000 cameras operated primarily for the benefit of law enforcement agencies will be governed by rules that take seriously the interests of the people those agencies have power over. Right now, the answer is no. The cameras watch. The data accumulates. And for the dozens of people tracked by officers who had no business watching them, the system worked exactly as designed — just not for them.
Legislatures considering surveillance technology ordinances, federal lawmakers debating data privacy frameworks, and the communities currently weighing whether to sign Flock contracts have a concrete body of evidence now. The misuse was not hypothetical. It happened, at scale, across multiple states and departments. The question is whether that evidence will be enough to build the oversight architecture that should have existed before the first camera went up — or whether the next batch of cases will simply add to a record that nobody acted on.
