Connecticut has the highest per capita rate of automated license plate reader (ALPR) cameras in New England and across other northeastern states, including New York, Pennsylvania, and New Jersey.

According to DeFlock, Connecticut ranks twenty-fifth overall for the number of per capita cameras in the state, averaging roughly 1.78 cameras per 10,000 residents. 

Those cameras are maintained by a variety of police departments and private businesses and use artificial intelligence to take pictures of and ‘read’ the license plate and make and model characteristics of every vehicle that goes past. The cameras can be either mounted to roadside infrastructure, like a pole or traffic light, or in the dashboard of a police car.

Every plate read is dumped into a massive database law enforcement can search and check against various crime hotlists. The majority of the data collected and stored is not connected to a crime and is never searched for or viewed by police before it’s purged from the system, often after a 30-day retention period. 

To law enforcement, these facts show that ALPRs are not a privacy threat: ALPR cameras do not record images of drivers or people (though other cameras manufactured by the same vendors do have that capability), don’t record personal information, and are searched retroactively by police in connection with a specific crime or investigation. As ALPRs have increasingly become a topic of public concern, law enforcement officials have also begun to adopt policies regulating use of the cameras, sometimes in compliance with laws.

But to privacy experts, none of these apparent safeguards are enough to allay concerns that ALPR cameras enable constant, warrantless surveillance of the everyday movements of private citizens. That some municipalities, including in Connecticut, have installed ALPR cameras without public input or oversight does not help, nor do high-profile headlines revealing how external law enforcement, including federal immigration officials, have accessed municipal data.

Police departments often share access to data with external departments, sometimes enabling police from across the country to search data at will. While there are ostensibly policies in place to prevent this from being abused, such as requiring searches to be connected to an investigation, they are easy to get around. Audit logs of ALPR data often list vague keywords like “investigation” as the reason behind a search, making it impossible to know whether the access was legitimate or not.

ALPR data is also not easy to obtain. Though public-facing transparency portals provide a veneer of accountability, not all departments use them, nor are they always accurate. Information contained in audit logs is often redacted or meaningless. The result is that much information about ALPRs, what they capture, and how they are used is beyond the public’s ability to know.

ALPRs in CT

Public unease about the use of ALPRs is likely not helped by the relative secrecy with which the technology has been rolled out. While public reporting in the past few years has helped to dramatically increase the public’s understanding of the technology, ALPR cameras have actually been in use for decades and communities that have been using them have not necessarily publicized that fact. 

In 2024, for example, controversy erupted in Colchester after police began testing Flock Safety ALPR cameras under an unsigned memorandum of understanding (MOU) that the town’s Board of Selectmen never approved.

While the MOU did not allow for the installation of any cameras, a proposed agreement between the Colchester Police Department, which is supported by the Connecticut State Police’s resident trooper program, and Flock Safety proposed installing three cameras in town. Though public pushback put a halt to the program, Colchester police still have access to ALPR data from at least one other police department in the state, according to publicly available data from Flock. 

Because so much surrounding ALPR cameras is hidden, it’s difficult to know exactly how many are in operation in Connecticut. DeFlock has mapped 639 cameras across the state, not all of which are maintained by police departments, and over thirty municipal law enforcement agencies who have contracts with ALPR vendors.

But it’s likely there are more. While DeFlock currently has the most comprehensive public data on ALPRs available, driven by public reporting of camera locations and public records requests, it is by no means complete. 

While town officials who use Flock vendors can use a ‘public transparency portal’ that reports basic information like the number of vehicles recorded in the past 30 days, the number of searches of retained data, and the number of vehicles flagged on various crime hotlists, many choose not to. Some also choose not to report the number of cameras they operate or the law enforcement agencies with whom they have agreements to share data, enabling police across the country to search municipal Connecticut data at will.

Using FOIA requests and relying on public reporting resources, such as DeFlock and Flock transparency portals, Inside Investigator was able to trace 185 cameras currently in operation around the state, mostly in towns that use Flock as a vendor. 

While FOIA requests Inside Investigator sent to a number of police departments seeking ALPR vendor contracts were still pending as of publication (several months after they were submitted), that makes the number of cameras traceable using public data a mere fraction of those that, according to DeFlock, are actually in operation in the state. 

Documents obtainable through FOIA also don’t necessarily reveal everything the public might want to know about ALPR cameras, such as the location of cameras or the direction they’re pointed. 

But ALPR camera location has also been a source of public concern, with cameras in some states pointed at sensitive locations like healthcare facilities, lawyers’ officers, or even in a handful of cases directly at private residential driveways, making it impossible for people to come and go from their own homes without their movements being recorded.

In Cleburne County, Georgia, a Flock ALPR camera was pointed directly at one family’s residential driveway, capturing images of the vehicles of anyone who entered or left the house. 

The Institute for Justice (IJ), a nonprofit public interest law firm, has brought several lawsuits challenging the constitutionality of ALPR cameras based in part on the location and density of ALPR cameras. 

In one of those cases, a federal class action lawsuit being brought against the city of San Jose, California and its police department, IJ attorneys allege their client Tony Tan is unable to leave his home without the government knowing due to the density of the city’s 474 ALPR cameras, which are strung along the streets Kann must take to leave or enter his apartment complex. The lawsuit expresses concern that Tan, who is politically active and monitors Immigration and Customs Enforcement (ICE) activity, could be monitored and targeted for his political speech as a result of the cameras’ ability to track his movements. In the wake of protests over amped up immigration enforcement raids, ICE has been accused of using license plate data to track legal observers and protesters exercising their First Amendment rights.

Michael Soyfer, an IJ attorney representing Tan and others in the case, told Inside Investigator that they have also identified blocks where potentially sensitive locations like churches, immigration attorneys’ offices, and healthcare facilities were encircled by ALPR cameras, enabling law enforcement—including any police across the country with whom San Jose shares data—to track visitors to those locations.

Connecticut does not have the same camera density, but Inside Investigator mapped known camera locations against potentially sensitive locations like healthcare facilities, immigration clinics and courthouses and found a number of locations where ALPR cameras are positioned along roads that lead to Planned Parenthood clinics, hospitals, and immigration legal clinics, in some cases making it possible to track anyone who comes and goes due to the road layout.

Other locations may seem more innocuous, targeting busy intersections and shopping plazas, but again make it difficult for anyone to come and go, even on routine shopping trips, without having their movements recorded.

Legal Framework

Based on ALPR camera data Inside Investigator was able to obtain through FOIA, a single license plate can be captured thousands of times by ALPR cameras over the course of a month. 

But is that cause for concern? Law enforcement and ALPR vendors will certainly say no, pointing to the cameras as a low-cost, low-manpower tool that’s useful for solving crime. They’ll also argue, as the Manchester, Connecticut Police Department did in a Freedom of Information Commission complaint Inside Investigator brought, that they are unable to search the vast majority of data that’s collected, implying it’s not a privacy concern.

A number of civil liberty and privacy advocates disagree that what is inarguably mass data collection and retention is so innocuous.

The lawsuit against San Jose is the second IJ has brought challenging the constitutionality of ALPR cameras. In its first case, IJ represented Lee Schmidt and Crystal Arrington of Norfolk, Virginia and argued that the capacity of the city’s 172 cameras to track people’s movements over Flock’s standard 30-day retention period of ALPR camera data represented a search requiring a warrant under the Fourth Amendment.

The lawsuit noted that not only do Flock cameras effectively track those driving in the city by capturing and storing their license plate and location data, but that anyone with Flock login credentials can access and search that data once stored.

“All of this is done without a warrant. No officer ever has to establish probable cause, swear to the facts in a warrant application, and await the approval of a neutral judge.” IJ’s complaint in the lawsuit states.

Soyfer said the case first came to IJ’s attention when the Norfolk police chief said it would be “difficult to drive anywhere of any distance without running into a camera somewhere.” He told Inside Investigator that ALPR cameras from vendors like Flock are creating a “real world panopticon” where law enforcement knows where everyone is going and when and that they can access to “mine at will.”

IJ’s lawsuits, part of its Plate Privacy Project, are encouraging the courts to “return to the history of the Fourth Amendment” according to Soyfer.

In previous decisions, Soyfer said the Supreme Court has said that individuals have a reasonable expectation of privacy in their long term physical movements, and while there’s a general idea that people can see you if you’re out in public and potentially follow you for a short time, figuring that out over a long period of time has been “resource intensive and traditionally hasn’t been feasible.”

ALPRs, IJ is arguing, change that feasibility and make city-wide surveillance possible, but are “inconsistent with the historical values that underpin the Fourth Amendment.”

U.S. District Court for the Eastern District of Virginia, Norfolk Division, judge Mark Davis ruled against IJ, finding ALPR cameras do not violate the Fourth Amendment, but the case is currently on appeal.

Soyfer said Davis’ finding created a “really impossible standard” where “unless you get zero gaps in people’s movements, your actions aren’t subject to the Fourth Amendment.” Soyfer believes the appeal has a strong chance of succeeding because it’s being heard by a court that struck down a Baltimore drone program that took second-by-second photos of the city and could trace individuals as blurred dots.

“It’s much easier when you have a unique identifier like a license plate where you can link it across time, which is assisted by AI.” Soyfer added.

Proponents of ALPRs sometimes point to the fact that cameras are usually found on and pointed at public roadways and that individuals generally don’t have an expectation of privacy in public.

But, according to Soyfer, that argument doesn’t erase Fourth Amendment concerns about the data ALPRs collect and store.

“There’s something qualitatively different between isolated observations of people in public at specific points of time and a record of where they were over the past 30 days, etc, in a government database that’s easily searchable and can be reconstructed.” Soyfer said. “Those are not subject to public observation, unless someone is stalking you or the police are putting you under surveillance.”

Soyfer added that argument “really misses the forest for the trees, because what people are objecting to is not what any one camera sees at any one point in time, it’s the creation of city-wide, state-wide, nation-wide systems that are creating huge troves of data without any admission they’ve done anything wrong.”

IJ’s lawsuit in San Jose also argues that the lack of cameras in saturated areas matters too, allowing law enforcement to make inferences about where people are not travelling.

“By identifying where a car was not captured, officers can home in on its route and exclude areas where the driver did notgo. In other words, the [San Jose Police Department’s] decision to strategically position hundreds of Flock Cameras throughout San Jose means that [they] can often deduce where a person could have gone even after their car left a given camera’s field of vision or during intervals when the person’s car was not spotted by any cameras at all.” the complaint states.

In heavily trafficked intersections, many cities that use ALPR will position multiple cameras facing in different directions, enabling police to determine not only when a vehicle passed through but in what direction they were traveling. Cameras in Connecticut follow this pattern.

“With two or more captures on a given route, [San Jose Police Department] officers can connect the dots and infer where someone went. This is helped by the fact that officers can also easily look up the person’s home address with the license plate numbers and use public records to find out more information about them.” the complaint continues.

As the lawsuit notes, in response to public concern, San Jose’s City Council did adopt a resolution prohibiting Flock cameras near reproductive healthcare facilities that perform abortions and places of worship. However, the police department relied on a narrow definition of when a camera was near one of those locations to avoid moving any of them.

“Coordinated attack” 

Within the past year, as immigration raids have filled state and national headlines, concern about the proximity of ALPR cameras to sensitive locations like reproductive healthcare facilities and immigration clinics has gained increasing political momentum. Connecticut, like a number of other states, introduced a number of bills to limit ALPR use and information sharing after news reports showed that data collected by the cameras had been accessed by immigration enforcement officials.

Part of a broader bill limiting how federal immigration officers can operate in the state, the ALPR provisions within SB 397, which has been enacted and will go into effect on October 1, include:

• Limiting the retention period for data captured by ALPR cameras to 21 days unless certain conditions, such as use in an active criminal investigation, are met
• Prohibiting public agencies from entering into contracts with ALPR vendors or using ALPR data in connection with investigations based on immigration enforcement, speech protected by the First Amendment, protected categories like race and sexual orientation, or prosecuting anyone seeking reproductive healthcare services
• Limiting law enforcement use of ALPR systems to comparing data against various state and federal crime hotlists
• Prohibiting sharing of ALPR data with law enforcement agencies outside of the state, except for in New York, Rhode Island, Massachusetts or as part of a multijurisdictional task force, or if a law enforcement agency requesting access has a warrant

The law also creates requirements for law enforcement agencies to create written system usage and privacy policies and directs the Police Officers Standards and Training Council (POSTC) to create privacy guidelines and, in concert with the Department of Emergency Services and Public Protection and the Institute for Municipal and Regional Policy at the University of Connecticut, to create a standardized form for publicly reporting ALPR camera usage.

POSTC recently released a model policy for ALPR usage, which includes standard policies for data management, access, training, and managing equipment and meets several of SB 397’s requirements. For example, it prohibits accessing ALPR data for investigations related to immigration enforcement or reproductive healthcare and states data can only be queried for legitimate reasons related to an investigation. It also requires users who are querying data to enter a reason for the query, such as an offense type, and case number.

According to data reported on Flock transparency portals, a number of police departments already follow either POSTC’s model policy or other model policies put forward by law enforcement agencies, such as the Commission on Accreditation for Law Enforcement Agencies.

But determining whether those policies are being followed is not so simple—and a provision in SB 397 exempting ALPR data from FOIA disclosure doesn’t make public oversight of law enforcement use of ALPR technology any easier.

Most ALPR vendors produce two types of audit logs that allow the public to see when data has been searched and accessed by police: organization audit logs, which contain information about searches performed internally by an agency, and network audit logs, which contain information about searches conducted by external agencies to whom a police department has granted access to their data.

Ostensibly, those logs, some of which are available through Flock portals, allow the public to understand whether ALPR data is being accessed for legitimate purposes and by whom.

But, in reality, publicly available data shows that, despite existing requirements that searches of ALPR data contain reasons for searches, that’s not always the case. Or, if they do, that reason is often a vague word like “file” or “investigation” that doesn’t really reveal whether information has been accessed for a legitimate purpose.

Flock’s platform allows police to enter either a reason for a search or an offense type. Data Inside Investigator reviewed, obtained both through FOIA requests and via some department’s Flock portals, usually contained either a reason for a search or an offense type, but rarely both.

In data we reviewed, not all recorded searches contained a reason. For example, in data from Cheshire police, no offense was listed in 37 of the 516 searches of the department’s organization audit logs from the past 30 days.

The specifics of other reasons for searches varied. Some searches listed both a specific crime and location. For example, “Cheshire shoplifting” appeared eight times as a search reason. “Erratic” was entered as a reason in 390 searches.

Other searches were more cryptic. “File 1” appeared in a number of searches, not only in Cheshire’s data but in other departments’, sometimes in combination with other descriptors like a location. Other variations, like “F1” or “F16”  or “file 16” appeared in a number of searches, as did strings of numbers. “Investigation” also appeared frequently as a reason for search.

For those organizations that reported an offense type in relation to a search, more detailed information about the crime being investigated was consistently available, but that was not always reported by departments.

Nor do all police departments make the name and badge numbers of officers who have conducted searches available, again making it harder to determine whether ALPR data is being accessed in accordance with policy.

In December 2025, Flock also made changes to what kind of data is included in network audit logs, automatically removing officer names, plate numbers, and vehicle fingerprint information. 

“Flock was founded on the principle of promoting public safety, while partnering closely with law enforcement agencies across the country to bring clarity and allow accountability, which helps to maintain public trust.” the company wrote in an email apparently distributed to all law enforcement officers with login credentials. “That commitment to transparency remains unchanged, but the environment around us has changed dramatically. We cannot let those abusing our transparency compromise officer safety and the safety of our community.”

As a result, network audit logs now hide information about officers accessing departmental information even from the agency that collected the data. In the same email, Flock suggested any agency that had concerns could reach out to an agency that searched their data for more information.

In the same email, they announced they were requiring searches to enter an offense type, in an effort to “eliminate subjective or inconsistent entries while strengthening compliance and public accountability.” Not all data Inside Investigator reviewed contained this information.

But from a public perspective, the automatic redaction of officer names means that, once again, little information that would promote oversight is available. Obtainable through FOI, network audit logs contain the name of an organization whose employee searched departmental data, as well as the time frame for the search and the total number of times that the network has searched departmental data in the last 30 days. But the names of individual officers and the filters used in a search are redacted.

In data Inside Investigator reviewed, searches originating in organizations outside the police department who had collected the data vastly outnumbered searches by departments.

For example, between April 6, 2026, and May 6, 2026, Cheshire Police Department officers searched their ALPR data approximately 1,680 times according to departmental audit logs. According to network audit logs, external agencies searched Cheshire data just under 19,000 times.

In Monroe, between April 5, 2026, and May 5, 2026, police officers searched data collected by their cameras 58 times. External agencies searched Monroe ALPR data 11,493 times.

Even before ALPR cameras came under scrutiny for their role in aiding immigration enforcement activities, a number of headlines, largely reported by 404 Media, revealed security lapses and concerning practices from Flock. In December 2025, a security vulnerability left the livestream of approximately 60 of its Condor cameras open to the general internet.

Unlike the Falcon ALPR cameras, Condor cameras are able to track individual people in real time. The vulnerability allowed anyone to access camera administrator control panels, watch streams from the cameras, download video archives, run diagnostics, and change settings. 

Another story from 404 Media exposed Flock’s use of Upwork gig workers located in the Philippines to train its camera’s machine learning algorithms. The identities of millions of surveillance targets were also unmasked after unredacted Flock audit logs were publicly released in January 2026. In response, Flock threatened the creator of HaveIBeenFlocked, a website that turned the leaked data into a searchable database, with a lawsuit claiming he violated their intellectual property rights.

Flock has generally not reacted well to media scrutiny, attributing negative headlines to the coordinated efforts of anti-police activists and accusing websites like HaveIBeenFlocked and DeFlock, which are collecting and organizing data that’s been made available through public records requests, of exploiting FOIA laws.

Following the leak, Flock utilized the direct contact it has with police departments to send an email, apparently to every law enforcement officer with login credentials, saying that Flock and its law enforcement clients were “under coordinated attack.”

“The attacks aren’t new. You’ve been dealing with this forever, and we’ve been dealing with this since our founding, from the same activist groups who want to defund the police, weaken public safety, and normalize lawlessness. Now, they’re producing YouTube videos with misleading headlines. They’re also trying to turn a public records process into a weapon against you and against us.” Flock CEO Garrett Langley wrote in the December 8, 2025 email.

More recently, Flock has argued that it is more transparent than other ALPR vendors because it makes audit logs—albeit with mandatory redactions of information—available through records requests.

14 FOIAs

The suggestion in Langley’s email, that public records requests for ALPR data are being weaponized against law enforcement, seems to be an attitude that permeates quite a few police departments in Connecticut.

For the past two years, Inside Investigator has been attempting to collect data from ALPR cameras, as well as other records that document exactly what types of data ALPR cameras collect and how it is used by law enforcement.  

Despite the spin, these efforts are not an attempt to undermine police powers and are rooted in clearly legitimate journalistic purposes, chiefly the idea that each person has a right to know how a technology capable of tracking their movements over an extended period of time is being used.

To be clear: some departments have fully and quickly complied with FOIA requests, both for ALPR data and for audit logs. Others have been less forthcoming, either claiming the data is exempt because releasing it would reveal investigatory law enforcement techniques and jeopardize police investigations, or simply ignoring requests entirely.

In November 2024, Inside Investigator first filed a series of FOIA requests with 14 different state and municipal police departments that, according to EFF’s Atlas of Surveillance, had contracts for ALPR cameras. Those requests sought 30 days of data captured by cameras.

Just three departments complied with the request and turned over data: Hartford provided data from its Rekor and Axon cameras, the University of Connecticut from its ALPR system, and Southington provided not ALPR data but a report showing the number of cars that had been captured by cameras.

 The New London Police Department provided a completely redacted image of its ALPR database—and never cited an exemption justifying withholding the data.

Four other departments denied the request, all on different grounds. The New Haven Police Department argued that there were no responsive records because ALPR data is uploaded into a database, which must then be manually downloaded to create a record, which it claimed did not fit the definition of a record under FOIA.

The Norwalk Police Department cited FOIA’s exemption for “any investigatory techniques not otherwise known to the public.” 

The Cheshire Police Department cited the law enforcement exemption, an exemption for documents related to collective bargaining, “and possibly others” and did not respond to requests for clarification. 

The Manchester Police Department cited the exemption for records that are compiled in connection with the investigation of a crime.

Six police departments—New Canaan, Danbury, East Haven, Greenwich, Stamford, and the Department of Emergency Services and Public Protection (DESPP)—s imply never responded to the requests, which are outstanding to this day.

The ALPR data Inside Investigator was seeking does not reveal information about specific crimes, police investigatory techniques, and, as proponents of ALPR technology frequently argue, was not collected in connection with a crime.

As data Inside Investigator obtained from the three police departments that did turn over records shows, ALPR data produces a database, where each line of data represents an individual capture of a license plate and vehicle information by a camera.

How that is used by law enforcement is not recorded in that data—though it is recorded in other data, such as audit logs, which are far easier to obtain.

As a result, and believing that there is public interest in the disclosure of ALPR data, Inside Investigator decided to challenge the Manchester Police Department’s (MPD) denial because it seemed easily demonstrably not connected to the underlying records. With an FOIC ruling stating the exemption did apply, we could go back to the other departments that had denied or ignored our requests and renew them.

Long story short, the FOIC agreed the exemption likely did not apply, but there were complicating factors. Once a request for documents is submitted through FOIA, public agencies are supposed to preserve those documents. MPD, which uses Flock as a vendor, did not do this and allowed the 30 days of data we requested to be automatically purged from their system. 

MPD testified at a hearing on the complaint, in which they raised for the first time an argument that the request was too broad and vague to comply with, that they could not even access all the data they collect in a day due to its size.

They presented no evidence that they talked to Flock about preserving the data we requested, for which they were chastised by the FOIC. In its final decision, the FOIC found that the data at issue was likely not exempt from FOIA disclosure. However, because MPD had purged the data and the FOIC was not able to review it, they could not issue a definitive decision.

They gave us the ability to file an expedited appeal if MPD denied a follow-up request for the same data—which is exactly what happened.

Seventeen months after Inside Investigator filed an initial request for 30 days of ALPR data, we again filed a nearly identical request, this time specifying that the 30-day time frame began whenever MPD downloaded the data in an attempt to prevent MPD from being able to work around the request by automatically purging it.

MPD denied that request on largely the same grounds as our previous request and despite the FOIC’s findings. This time, MPD attempted to argue that it could not comply with the request for data because it is maintained by Flock, cannot be accessed by MPD without Flock’s help, and is therefore not a public record.

This time, MPD cited additional parts of FOIA’s law enforcement exemption, including the investigatory techniques exemption the FOIC already largely repudiated, and another portion of the exemption for “information to be used in a prospective law enforcement action if prejudicial to such action.”

The morning of the hearing in the appeal, MPD also attempted to have it delayed so that officials could begin the process of contacting DESPP to make a safety determination about the release of the requested records. FOIA has a process that allows public agencies to seek a determination about whether the release of certain types of records would create a safety risk from certain legal agencies.

MPD also cited what at that time was pending legislation to exempt ALPR data from FOIA as justification to delay the hearing. The hearing officer denied the request, however, the case is still pending before the FOIC.

Pulling Back the Veil

To date, Inside Investigator has not been able to obtain Flock ALPR data. The data we have obtained from other systems, however, sheds light on how ALPR cameras operate.

While there are differences across systems, they generally all record:

• The location of the camera
• The time of the capture
• The license plate of the vehicle
• Vehicle characteristics like color, make, and model

Some systems also capture things like whether a vehicle has bumper stickers and where they are located. ALPR uses AI to read most of this data and sometimes contains a percent probability of its “confidence” in the accuracy of the read. Weather conditions, light quality and camera calibration all can impact the camera’s “confidence” of its capture—and errors in how plates are read and recorded have led to wrongful detainments and arrests when an ALPR camera incorrectly read a plate number.

Hartford’s Axon data, for example, includes confidence stores for the type of license plate, vehicle make, model, color, whether there were window or bumper stickers, whether a vehicle had top or back racks, and vehicle type. Confidence for those characteristics varied greatly, ranging from near 100 percent to zero, sometimes in the same vehicle read.

For example, for one vehicle capture that occurred on October 28, 2024, the camera was 99 percent confident of the plate read as a regular Connecticut plate, was unable to determine the vehicle color with 49 percent confidence, was 48 percent confident the vehicle was a Nissan, 72 percent confident it was a sedan, and less than one percent confident the vehicle had bumper stickers.

ALPR data also reveals just how frequently vehicles can be captured by cameras, particularly if they frequent areas near where cameras are located. In the data Inside Investigator obtained from Hartford, individual license plates were recorded up to hundreds and sometimes thousands of times over the course of 30-days, sometimes with multiple plate reads captured within minutes of each other.What this data does is help shed light on how ALPRs operate, what type of data they’re collecting, and help the public understand how their movements might be captured by ALPR cameras.

What it does not do is reveal anything about police investigations or how ALPRs are used as an investigatory technique.

But there are other records that do, which are much less frequently subject to challenge in public records releases.

Aside from the claim that releasing ALPR data to the public reveals police investigatory techniques, law enforcement has also claimed that license plate data is potentially sensitive. It’s an argument that drove exemptions for ALPR data from public records laws in a number of state legislatures this year.

License plate numbers by themselves are not private data—anyone can go into a parking lot and write down the plate numbers of cars parked there. Most states, Connecticut included, also exempt personally identifying information tied to license plate numbers from public records disclosure. But add a record of where and when a vehicle was, and some privacy experts and law enforcement entities worry that disclosure of the data has the potential to jeopardize the safety of certain groups, such as domestic violence victims.

Yet, license plate data, as well as other potentially sensitive information like police investigation details that are generally exempt from FOIA, is present in other ALPR-related documents that are subject to public disclosure, and which numerous police departments turned over without objection.

Numerous audit logs Inside Investigator received from police departments in response to a series of FOIA requests, for example, contain license plate numbers (Inside Investigator has redacted this in all data we are publishing in order to protect personal privacy) and, in some cases, where a reason for a search was entered, specific investigation numbers. 

Requirements to enter reasons for searches reveal details about the types of crime being investigated, sometimes also providing information like towns, street names, house numbers, or when investigations involve multiple police departments. All of that helps provide more detailed information to the public about how ALPR technology is used by law enforcement, which serves a legitimate public interest.

Requests for communications between police departments and their ALPR vendors also may contain information about law enforcement investigations–and help shed light on how ALPRs are used and how police interact with law enforcement vendors. Flock and other ALPR vendors are private companies, which means their records are largely hidden from the public—their interactions with law enforcement vendors, disclosable through public records, are one of the few means the public has to gain a greater understanding of how the technology is used.

In a records request Inside Investigator submitted to MPD, seeking communications between their officers and Flock—in an attempt to determine whether they had contacted Flock to comply with our FOI requests—emails helped reveal:

• How police communicate with Flock about camera calibration issues
• How requests for information sharing between different police departments work
• How Flock’s systems flags captures of vehicles that are wanted by multiple jurisdictions
• How permitting of cameras and issues with camera locations are handled
• How Flock solicits additional business from its clients with email marketing and direct outreach from its employees
• How Flock uses its direct connection to its clients to respond to headlines criticizing its products and market its successes
  

FOI Exemption

There are still holes in public knowledge about what data ALPRs collect and how they are used. And, with ALPR data set to shortly become exempt from FOIA, those holes will not get any smaller. In Connecticut, at least, the FOIA exemption for ALPR data that passed as part of SB 397 was advanced on the idea that, even though widespread data collection is justified because of the role it can play in solving crimes, allowing it to be publicly disclosable presents a threat to personal privacy.

But privacy experts question whether that’s true.

“It hasn’t been clear to me that there has been any harm done through the use of ALPR information, except when it comes to more oversight and identifying cases where ALPR data and information has been used inappropriately.” Beryl Lipton, a senior investigative researcher at the Electronic Frontier Foundation (EFF), told Inside Investigator.

“There are already exemptions in public records laws that would allow [public officials] to exempt dangerous information.” Lipton added.

EFF, a nonprofit that advocates for digital privacy and free speech, has tracked the growth of ALPRs across the country, including through its Atlas of Surveillance, and documented numerous abuses of the technology, including audit log search terms that contain racial epithets and searches of ALPR data connected to protestors whose speech is protected by the First Amendment. Like journalists around the country—including Connecticut journalists who found municipal ALPR data was searched by out-of-state immigration officials—EFF primarily relied on public records requests to uncover these abuses.

But, as Lipton noted, there’s a certain irony to laws like SB 397, the majority of which was aimed at curbing these abuses of ALPR cameras, and that brought to lawmakers’ attention by reporting that will likely become more difficult once the FOIA exemption goes into effect.

“You’re seeing some of these exemptions, in Connecticut and elsewhere, are being introduced in bills that contain other nods to public transparency, and I think it’s a very intentional packaging to try to get some of these exemptions through, knowing the public has found great value in being able to access that information.” Lipton said.

SB 397’s FOIA exemption is for ALPR data, defined as “any data captured, recorded, held or stored, processed or derived from an automated license plate reader system, including, but not limited to, license plate characters, vehicle still or video images, vehicle attributes, location data, time stamps and metadata.” While the law does specify that camera locations and audit logs are still disclosable, and that individuals may request ALPR data for their own vehicles, the exemption is still broad—and likely to mean that data currently reportable in audit logs will be redacted.

“It exempted information in such a broad way. There are privacy reasons why, if I were to ask for all of a particular person’s scans, that under the existing privacy exemptions in most public records laws, that information already could be excluded. You could also be able to make the argument that you also couldn’t see who agencies are sharing information with.” Lipton said.

And that last point is of particular concern. That law enforcement agencies across the country, at all levels of government, are able to access municipal ALPR data is already broadly known—and is an area of broad concern.

However, there’s a level of information sharing that goes beyond cooperation between agencies when it comes to routine crime solving, which is what frequently gets touted by those who advocate for ALPRs.

“ALPR data doesn’t just get collected and stored, it goes into a data sharing network, and increasingly it’s going into other networks designed to share and analyze information, much more like how phone numbers get spread around than a point and shoot camera taking pictures of cars. We really don’t know where it’s going and we’re not going to know.” Lipton said of the larger impacts of the FOIA exemption for ALPR exemptions.

“It’s most likely [data] is also going to be going into real time crime centers, which is designed to ingest all information and apply AI to try to generate so-called leads.” She added. “By exempting all of the information in the way that [Connecticut has], it’s substantially harder to understand, once you get beyond the initial privacy information, now not to be able to understand all the additional uses that raw data is going to be implicated in.”

Lipton also noted that SB 397 included several measures aimed at providing transparency around data brokers and regulating the markets on which they share and sell personal information.

“It’s a weird coupling if the principle of what you’re trying to do is provide protection.” Lipton said of brokering appearing in the same bill as the FOIA exemption for ALPR data.

Even outside of FOIA, some of the apparent steps ALPR vendors like Flock have taken to increase transparency can actually make understanding ALPR use more difficult. 

Flock, and a number of towns who contract with them, point to their transparency portals as a tool the public can use to understand how ALPRs are being used in their towns. But is the data posted there trustworthy? 

That can be difficult to determine, especially when police departments who have those portals point to them to fulfill FOIA requests rather than turn over underlying records, like audit logs, that would help confirm the veracity of what’s posted online. In response to a request for audit logs submitted to numerous police departments across the state in May 2026, several police departments pointed to their public portal rather than turn over underlying documents, including the Milford Police Department and the Putnam Police Department.

Lipton said EFF has found cases where police departments have not listed all of the agencies with which they share information on their public portals.

“You’re doing more obscuring reality than providing transparency, and providing this idea that everything you see on the portals is accurate and comprehensive when actually it is not.” Lipton said. 

The FOIA exemption is also of concern because Flock and other ALPR vendors who are collecting data in bulk are private companies. FOIA requests for documents that capture interactions between police and those vendors are really the only avenue the public has to understand how ALPR companies operate. 

Another problem: there’s a fundamental imbalance between the access that ALPR vendors selling their product to police have and the access that the public and others concerned about privacy have.

“It speaks to a problem we have in policing right now, where law enforcement has a much closer relationship with private vendors who want technology to be adopted than do the public. We’re not putting principles in place to protect the public when we’re privileging the arguments of people to implement these technologies.” Lipton said.

And exempting ALPR information from public view may make that imbalance greater.

“Police trust vendors are only telling part of the story. We need journalists and the public to be part of the conversation, to be really getting into these details because this is mass surveillance.” Lipton said.

ALPRs Town-by-Town

Right now, there is no single database in the state that lists every town and agency either using ALPR cameras or that has access to data collected by them. It’s possible the public reporting requirements in SB 397 will force more of this data into the public. It’s also possible the FOIA exemption in the law will make it more difficult to independently collect and verify publicly reported data.

Inside Investigator has filed dozens of FOIA requests in an attempt to gather as much information as possible. Most recently, we filed a series of requests for ALPR vendor contracts and audit logs to over 30 police departments across the state. At time of publication, more than half of them are outstanding over 60 days later.

View all ALPR data and contracts we’ve obtained to date in our FOIA Library.

In our FOIA library you can find a town-by-town breakdown of the data we were able to obtain, including ALPR vendors, the number of cameras located in municipalities, information about the number of cars collected, and how data is accessed by local police departments and external agencies.

The Wilton Police Department did turn over audit logs, but provided PDF screenshots of audit logs, which do not capture the full view of the dashboard and therefore could not be analyzed.

The Enfield Police Department does not have a direct contract with an ALPR vendor, but does have a bid waiver with Vulcan for its Joint Operations Center. They provided audit logs that only contained a list of user IDs and emails. “License plates are redacted due to the quantity.  It is not possible to identify which plates are part of ongoing investigations.” Enfield police wrote in response to Inside Investigator’s FOIA requests.

Windsor has paused its contract with Flock after public pushback.

All ALPR data Inside Investigator has obtained through FOIA requests as well as additional visualizations are available at our FOI Library.

Author’s Note: I would like to give special recognition to the staff and students at Yale’s Media Freedom and Information Access Clinic for their diligent work in representing me in my complaint against the Manchester Police Department before the Freedom of Information Commission.

Transparency Note: Author Katherine Revello testified in opposition to several bills proposing to exempt ALPR data from FOIA on behalf of the Connecticut Council on Freedom of Information.

Republish This Story

Republish our articles for free, online or in print, under a Creative Commons license.