After a Washington Post investigation found officers accused of using license-plate readers to stalk girlfriends, ex-wives and other private targets, Flock Safety is rolling out new safeguards across its vast surveillance network. The company says the changes will curb abuse. Critics say the bigger problem is that one private system now tracks too much of American life.
WASHINGTON - Filed at 4:01 p.m. ET
Flock Safety built one of the most powerful surveillance networks in America by making police work feel faster, cheaper and more searchable. Now the company is trying to prove that the same network can be made safe enough for the public to trust.
After a Washington Post investigation found that at least 50 law enforcement officers had been accused, charged or convicted of misusing Flock and other license-plate-reader systems, including cases in which officers allegedly tracked romantic partners, ex-wives or women without their knowledge, Flock announced a package of changes meant to limit abuse. The company says it will require more search accountability, shorten the default retention period for license-plate data, expand automated misuse detection and give agencies more precise control over how their data is shared.
The announcement is an important shift for a company that has often argued that local police departments, not Flock, control the rules for how the technology is used. But it is also an admission of something privacy advocates have been saying for years: when a private company builds a national system that can reveal where vehicles have been, voluntary safeguards are not enough.
The debate is no longer whether Flock cameras help solve crimes. They do. Police departments credit the system with helping find stolen cars, locate missing people, identify suspects and connect investigative leads across cities and states. For officers trying to solve a shooting, kidnapping, hit-and-run or carjacking, a plate hit at the right time can be the difference between a cold lead and an arrest. The harder question is whether that public-safety value can justify a network that captures the movements of ordinary drivers who are not suspected of anything.
Flock’s system is built around automated license plate readers, often mounted on poles along roads, near neighborhoods, outside businesses, at apartment complexes or in shopping areas. The cameras capture license plates and vehicle characteristics, then make that information searchable. A police officer can type in a plate and see where that vehicle has been spotted by the system during the retention period. Agencies can also share access with other departments, expanding a local camera into part of a much larger web.
That is what makes the tool valuable. It is also what makes it dangerous. A single camera can help identify a suspect’s car. Thousands of cameras can reconstruct patterns of movement. A search can reveal visits to a doctor, a church, a mosque, a political meeting, an immigration lawyer, an addiction-recovery group, a domestic-violence shelter or a former partner’s home. Even if the system does not show a person’s face, a license plate can be personal enough to expose a life.
The Post’s investigation made the concern concrete. The issue was not hypothetical misuse by some future bad actor. It was officers allegedly using systems designed for public safety to monitor private people. In several cases, the people being watched were women connected to the officers personally. That distinction matters because surveillance misuse is not only a civil-liberties issue. It can become a domestic-abuse tool.
That is the moral center of the controversy. A database that helps police follow a stolen car can also help a rogue officer follow an ex-girlfriend. A tool that finds a suspect can also reveal whether someone went to work, home, court, a date, a hospital or a shelter. The abuse does not require hacking. It requires access, curiosity and weak oversight.
Flock’s new rules are meant to tighten that gap. The company says law enforcement customers will be required to use an audit tool designed to detect abnormal search behavior. When suspicious activity is detected, the user can be locked out pending internal review. Officers will also be required to enter a code from their records-management system tying each search to a specific case before running a query. Emergency overrides will be allowed, but they will be automatically flagged for review.
Those changes could make casual abuse harder. They also make lying more difficult. An officer searching a girlfriend’s plate would now need to attach the search to a real case code or misuse an emergency override that creates an audit trail. That does not eliminate abuse, but it raises the friction.
Flock is also reducing its standard data-retention default from 30 days to seven days. That is significant because retention is the difference between a short investigative tool and a month-long map of movement. A shorter default limits how much old travel data remains available for broad searching. The company says data can still be preserved longer when tied to evidence and a case number, and agencies may be able to set longer periods through policy approvals.
That is where critics see the weakness. A default is not a law. If departments can extend retention, create exceptions or rely on “Evidence Mode” broadly, the practical effect may be limited. A seven-day default is a step toward data minimization, but it does not answer whether millions of innocent drivers should be scanned at all. It does not decide whether police need a warrant for certain searches. It does not solve the problem of uneven local rules.
Flock is also adding more granular sharing controls. Instead of sharing data with another agency on an all-or-nothing basis, departments can place limits on what outside agencies may search for. In theory, one department could allow searches for homicide investigations while blocking searches connected to immigration enforcement. That matters because public anger over Flock has not centered only on stalking cases. It has also focused on cross-jurisdiction sharing, federal access, abortion-related investigations and immigration fears. The company’s message is clear: local agencies should remain in control, but the platform will now make control more explicit.
The problem is that trust, once lost, is harder to rebuild than a software setting. Privacy groups argue that Flock’s reforms still leave police departments policing themselves. An audit tool may flag unusual behavior, but an agency must act on it. A case-code requirement may discourage abuse, but a dishonest officer can still enter false information. A seven-day default may shorten retention, but it can be changed. A sharing control may narrow access, but only if officials configure it properly and communities know what has been allowed.
That is the accountability gap. Flock is a private company, but the power it enables is governmental. Local officials sign contracts. Police search the database. Private cameras can feed public investigations. Homeowners associations and businesses may participate. Data can cross city lines. The average driver may never know which camera captured their plate, which agency stored it, which officer searched it, or which other department received access.
This is why the controversy has become politically unusual. Opposition to Flock no longer fits neatly into left or right categories. Civil libertarians, immigration advocates, abortion-rights groups, conservatives skeptical of government overreach, progressive activists, local privacy boards and some law-and-order officials are all asking versions of the same question: who watches the watcher?
The backlash is becoming visible at the local level. Some cities and counties have canceled contracts or paused deployments. Activists have organized campaigns against automated license plate readers. Local meetings once dominated by crime concerns now feature residents asking about retention periods, federal sharing, warrants, audits, data ownership and whether the police chief can explain every search.
That local pressure matters because Flock’s growth model depends on municipal acceptance. A national network is built one town, one county, one police department and one private customer at a time. If local governments begin seeing the cameras as political liabilities, the company’s strongest advantage, scale, becomes its greatest vulnerability.
Flock’s defenders argue that the criticism ignores the victims helped by the system. They say the cameras can locate abducted children, recover stolen vehicles and give police objective evidence without relying on biased eyewitness accounts. They also argue that roads are public and that courts have generally allowed license-plate scanning in public spaces. To them, the answer is not tearing down cameras. It is using them responsibly. There is merit in that argument.
Public safety tools should not be dismissed simply because they can be abused. Databases, body cameras, dispatch records, phone records and criminal-history systems can all be misused. The existence of misuse does not automatically prove the tool should disappear. But the scale of Flock’s network changes the analysis.
A stolen-car database is one thing. A nationwide searchable system recording billions of vehicle observations is another. The larger and easier the system becomes, the more severe the consequences of one bad search. A rogue officer does not need to follow someone physically. The database can do the first step for him. That is why platform design matters, but law matters more.
The strongest protections should not depend on whether Flock’s leadership chooses the right settings or whether a police administrator notices a flagged search. Legislatures should set clear rules. Search purposes should be limited. Retention should be short. Sensitive categories should be restricted. Abuse should carry criminal penalties. Audits should be independent. Public reports should disclose how often searches are run, how often data is shared, how many users are disciplined and whether federal agencies have access.
Communities should also know where cameras are located and who owns them. A surveillance network cannot claim public trust while remaining functionally invisible to the public. If cameras are scanning public roads, residents should not have to discover them through activists, leaked logs or investigative reporting. City councils should approve deployments openly. Contracts should be public. Police should explain use cases before the cameras go up, not after controversy erupts.
The Post’s reporting did what oversight should have done earlier. It identified a pattern of abuse, turned individual allegations into a national story and forced the company to respond. That does not mean Flock’s changes are meaningless. They may prevent real harm. A mandatory audit tool could catch an officer before a pattern continues for months. A case-code requirement could deter casual stalking. Shorter default retention could reduce the searchable footprint of ordinary drivers.
But these changes are not the end of the debate. They are the beginning of a more serious one. The question before cities now is not whether Flock can improve its software. It is whether local governments are willing to govern surveillance technology with the seriousness it deserves. That means asking what crimes justify a search, how long data should exist, who can access it, whether outside agencies can search it, what happens when an officer lies, how residents can challenge misuse, and whether the benefits outweigh the chilling effect of being constantly recorded.
The answer may differ by community. A city facing persistent vehicle thefts, kidnappings or gun violence may see the cameras as essential. Another may decide that the privacy cost is too high. But no community should make that choice passively because a vendor promises guardrails after national scrutiny. The public has reached a new stage in the surveillance debate.
It is no longer enough for a company to say its technology is useful. It must prove that usefulness does not become unchecked power. It is no longer enough for police to say the cameras solve crimes. They must prove that innocent people are not being turned into permanent data points. It is no longer enough for city officials to approve contracts quietly. They must explain what kind of public space they are creating.
Flock’s reforms acknowledge the problem. They do not settle it. A system that can help find the missing can also help stalk the vulnerable. A tool built for investigations can become a tool of intimidation. A camera network built city by city can become something no city fully controls.
That is the challenge now facing Flock, police departments and the communities deciding whether to keep the cameras on. The issue is not only whether the company can detect misuse after it happens. It is whether America is comfortable building a road network where every car becomes searchable first and protected later.
Reporting and sourcing transparency note: This article is based on current public reporting from The Washington Post, the Associated Press, The Guardian, The Verge, Wired, the ACLU, Flock Safety public materials and local reporting on recent camera backlash.
Privacy and public-safety information note: This article is for news and public-information purposes only. Laws governing automated license plate readers, public-record disclosures, data retention and police database misuse vary by jurisdiction and may change quickly.