Minnesota Privacy Project FAQ
What is an ALPR camera?
Automated License Plate Reader (ALPR) cameras are high-speed imaging systems that photograph and digitally process vehicle license plates in milliseconds. A single camera captures thousands of plates a minute in any weather, day or night, at highway speed. Using optical character recognition (OCR), they convert plate images into searchable, indexed text in real time. Unlike a police officer making a single observation, ALPR systems run continuously, silently logging every vehicle that passes within range, whether or not the driver is suspected of anything.
SOURCE: [EFF →]
What data does each ALPR capture per read?
Every ALPR capture logs the full license plate number, the date and time of the scan, the GPS location of the camera, and a photograph of the vehicle, often including make, model, color, and an image of the occupant area. This data is uploaded to a cloud database where it is indexed and retained, typically for weeks to months depending on the operator's policy. Police agencies with network access can then search across every other agency's stored reads, not just their own.
SOURCE: [ACLU →]
What is "Vehicle Fingerprint" and how does Flock market it?
Flock Safety markets a patented feature it trademarked as "Vehicle Fingerprint" that identifies vehicles even when a plate is unreadable, obscured, swapped, or covered by a paper tag. The system analyzes physical attributes such as roof racks, bumper stickers, dents, aftermarket wheels, and body style to build a persistent profile tied to the physical vehicle rather than its plate. A vehicle can then be tracked across Flock's network by appearance alone, which makes evading plate-based surveillance far less effective than it sounds.
SOURCE: [Wikipedia →]
What happens to your data after it is captured?
Flock Safety pools plate reads into its own nationwide sharing system, historically called TALON, built around a "National Lookup" feature that lets a participating agency search cameras run by thousands of other agencies at once. EFF has found a single query can reach more than 83,000 cameras nationwide. This is a separate ecosystem from Motorola/Vigilant's competing "LEARN" network. Retention varies by contract, commonly 30 days by default but often extended. Beyond policing, plate-read data has been documented flowing to insurers, repo companies, and data brokers.
SOURCE: [VICE →]
Have ALPR misreads led to wrongful stops?
Yes, and documented cases can be severe. In 2020, Aurora, Colorado police pulled over a woman and four children at gunpoint and forced them to lie face down on pavement after an ALPR system confused her Colorado plate with a stolen motorcycle's Montana plate, matching only the numbers, not the vehicle type or state. The city settled the resulting lawsuit for $1.9 million. Journalists have documented more than a dozen similar wrongful stops nationally, several involving guns drawn on children, almost always traced to an unverified plate match or an outdated hot list entry that no officer double-checked before acting on it.
SOURCE: [CBS News →]
How many ALPR cameras are deployed in Minnesota and nationally?
Flock Safety alone reported crossing 100,000 cameras in mid-2026, deployed across more than 5,000 communities and thousands of law enforcement agencies nationwide. In Minnesota, ALPR systems are operated by Minneapolis, St. Paul, Hennepin County, Washington County, and dozens of suburban and outstate departments, with more added yearly. Exact statewide counts are hard to verify because many vendor contracts are shielded from public records requests, and crowdsourced projects like deflock.me exist largely to fill that gap.
SOURCE: [Tech Times →]
Where are cameras typically placed?
Cameras concentrate at points of movement: highway on-ramps and off-ramps, major intersections, bridge chokepoints, and parking garage entrances. They are increasingly documented near politically and medically sensitive locations, including clinics, places of worship, and immigrant community centers. Placement decisions are made by law enforcement agencies, frequently without public notice, community input, or a city council vote of any kind. The crowdsourced deflock.me project exists specifically because these locations are otherwise hard for residents to find out on their own.
SOURCE: [deflock.me →]
How do I identify a Flock Safety camera?
Flock Safety cameras are compact, solar-powered units mounted on dedicated poles or street infrastructure at roughly 10 to 14 feet high. The housing is typically a gray or black rectangular box, about the size of a large book, with a single wide-angle lens facing traffic and a small solar panel on top. Units are often labeled "FLOCK SAFETY" and deployed in pairs facing opposite directions to capture both approaching and departing traffic. The deflock.me project maintains a crowdsourced, OpenStreetMap-based map of verified camera locations across Minnesota and nationally.
SOURCE: [deflock.me →]
How fast is the ALPR network growing, and who is paying for it?
Flock's deployed camera count has grown roughly twentyfold since 2020, crossing 100,000 units by mid-2026. A major driver is money: federal Homeland Security grants, DOJ JAG funds, and post-pandemic stimulus dollars have made per-camera cost a negligible barrier for most departments, letting agencies expand without a direct budget fight or public vote. In effect, federal grant programs are subsidizing the buildout of a privately owned surveillance network that residents rarely get to weigh in on before it goes live.
SOURCE: [Brennan Center →]
Who is Flock Safety?
Flock Safety was founded in 2017 by Garrett Langley in Atlanta and has become the dominant ALPR vendor for U.S. law enforcement, describing itself as a "public safety operating system." It operates on a recurring subscription model: agencies pay annual fees per camera, giving Flock a direct financial incentive to keep expanding camera counts. The company has raised roughly $950 million in venture funding at a valuation near $7.5 billion, with backers including Andreessen Horowitz and Tiger Global, and it markets directly to police departments and city staff rather than through public deliberation.
SOURCE: [Wikipedia →]
Who else is in this market?
Motorola Solutions acquired Vigilant Solutions in 2019, absorbing its historical ALPR database into a broader law enforcement surveillance platform built around a data-sharing network called LEARN. Axon, known for Tasers and body cameras, has also entered the ALPR market. Flock, Motorola/Vigilant, and Axon between them control most of the U.S. law enforcement ALPR market, and interlocking data-sharing agreements mean a plate read can end up reachable by agencies far outside whichever vendor a given city actually contracted with.
SOURCE: [EFF →]
How do city and county contracts work?
Most ALPR contracts are signed at the police department or city manager level, often without a city council vote or public notice. They are typically multi-year subscription agreements with automatic renewal clauses. Vendors routinely designate pricing, deployment maps, and even contract terms as proprietary, which can exempt them from public records requests depending on the state. Residents often learn about a deployment only through local journalism or advocacy research, sometimes years after the cameras went live.
SOURCE: [MuckRock / EFF →]
Do federal agencies access this data?
Yes, extensively. Reporting based on public search logs found more than 4,000 nationwide or statewide Flock lookups run by local and state police at the request of, or as a favor to, federal agencies, with officers logging search reasons like "immigration," "ICE," and "ICE+ERO." Separately, ICE has obtained access to Motorola/Vigilant's location database through a multimillion-dollar contract covering thousands of federal officers. Congress opened a formal investigation into Flock in 2025 after this reporting, but no federal law currently requires a warrant for this kind of purchased access.
SOURCE: [404 Media →]
Who buys the downstream data?
ALPR data amounts to location history, behavioral patterns, and inferred associations, and some operators and their partners treat it as a commercial asset. Public records requests and reporting have documented downstream access by insurance companies assessing driving behavior, repossession agencies locating vehicles, and private investigators, in addition to the thousands of law enforcement agencies with direct network access. In most states, no law specifically prohibits this kind of commercial use of plate-read location data, leaving it to vendor contracts and internal policy alone.
SOURCE: [ACLU →]
What is the track record on breaches and unauthorized access?
ALPR data has already leaked. In 2019, a CBP contractor, Perceptics, copied license plate and traveler images to its own network without authorization; that network was then hacked, and CBP confirmed the images of roughly 100,000 people were compromised. More recently, security researchers found Flock camera admin interfaces exposed without authentication, and audit logs containing millions of license plate searches have surfaced outside the access-control system meant to protect them. Centralizing this much location data in vendor-run systems creates a single point of failure that did not exist when the data was scattered and short-lived.
SOURCE: [TechCrunch →]
What is the public safety pitch, and does it hold up?
Law enforcement agencies and ALPR vendors argue the technology solves violent crimes, recovers stolen vehicles, and locates missing persons, and Flock claims its cameras assist thousands of arrests. Independent analysis has not shown statistically significant reductions in violent crime tied to ALPR deployment. Recovered stolen vehicles, the most frequently cited win, represent a small fraction of total reads, and catching them does not require logging every driver who passes by. The public-safety framing is politically effective and is often the only argument presented at the city council meeting where a contract gets approved.
SOURCE: [Brennan Center →]
How does ALPR route around the Fourth Amendment?
Courts have long held that a license plate visible on a public road carries no reasonable expectation of privacy, so a single ALPR read is likely constitutional on its own. The problem is aggregation: stitching together millions of individual reads produces a detailed movement history no single observation could. In 2018's Carpenter v. United States, the Supreme Court ruled that this kind of comprehensive location tracking is different enough to require a warrant when police compel a phone carrier's records. Most courts have not yet extended that reasoning to ALPR networks, leaving mass plate tracking in a legal gray zone Carpenter arguably should cover.
SOURCE: [EFF →]
How do FOIA exemptions shield private operators from accountability?
Because ALPR systems are run by private companies under contract to government agencies, system details can be withheld from public records requests under trade secret or proprietary-information exemptions. That can leave the public unable to determine how many cameras exist, where they're placed, what the retention policy is, who has database access, what the contract costs, or whether data is shared or sold onward. Private operation of what is functionally public surveillance infrastructure is increasingly used as a way to route around the accountability that would normally apply to a government program.
SOURCE: [MuckRock / EFF →]
How does commercial ALPR access bypass warrant requirements for police?
When a private company collects plate-read data and sells database access to law enforcement, agencies can simply subscribe rather than seek a warrant. This is the third-party doctrine applied at population scale: because a plate is exposed to a private company's camera on a public road, the government can buy access to that company's database without court authorization, even for searches that would plainly require a warrant if police ran the cameras themselves. Bills like the federal Fourth Amendment Is Not For Sale Act have targeted this exact loophole, but none has passed.
SOURCE: [ACLU →]
What does Minnesota law actually require, and where does it fall short?
Minnesota Statute 13.824 governs government-operated ALPR: data not tied to an active investigation must be destroyed within 60 days, agencies must undergo an independent biennial audit, and noncompliant agencies can be ordered to suspend use entirely. Washington County's ALPR program has gone through exactly this kind of state audit. The gap is that 13.824 regulates government-run cameras; it does not clearly govern private vendor networks like Flock's, where a Minnesota city buys access to a nationwide pool of data collected far outside any single agency's own retention or audit rules.
SOURCE: [MN Revisor's Office →]
What happened in 2025 that changed the conversation nationally, and are cities pulling back?
Reporting based on public search logs revealed a Johnson County, Texas sheriff's deputy ran a nationwide Flock search across more than 83,000 cameras, including in states where abortion is legal, logging the reason as related to a self-managed abortion. Separate reporting documented thousands of Flock lookups tied to federal immigration enforcement despite ICE having no formal contract with the company. Congress opened an investigation into Flock over both patterns, and cities responded: Evanston, Illinois terminated its contract after a state audit found illegal data sharing with federal agents, and by mid-2026 roughly 50 to 80 U.S. cities, including Austin, had canceled or rejected Flock contracts outright.
SOURCE: [EFF →]
How do I find out what ALPR data my city holds?
File a request under the Minnesota Government Data Practices Act (Minnesota Statutes chapter 13). Send it in writing to your city or county's "responsible authority," usually the city clerk or administrator, and ask specifically for any ALPR or license plate reader vendor contracts, the retention policy, the biennial audit required under section 13.824, and whether the system participates in any nationwide data-sharing network. Government data is presumptively public unless a specific statute says otherwise, so a vague denial is worth pushing back on.
SOURCE: [MN Data Practices Office →]
How can I help map cameras in my area?
The deflock.me project maintains a crowdsourced, open map of ALPR camera locations built on OpenStreetMap. Anyone can add a camera they spot, tagging its location and vendor, which is often the only way a location becomes publicly documented at all, since most cities won't release deployment maps voluntarily. Checking the map before a walk, drive, or protest route is also a practical way to understand your own exposure in the meantime, and every camera added makes the shared picture more complete for everyone else using it too.
SOURCE: [deflock.me →]
What should I actually say at a city council meeting?
Ask directly whether the city has a Flock, Motorola/Vigilant, Axon, or other ALPR contract, and if so, request the retention period, the audit schedule required under Minnesota Statute 13.824, and whether "national lookup" or similar network-sharing features are enabled. If no contract exists yet, ask for a public vote and a written data-sharing policy before one is signed, not after. Council members respond to specific, answerable questions asked on the record far more than to general privacy concerns, and a paper trail matters if the answer turns out to be wrong.
SOURCE: [ACLU-MN →]
How do I get involved beyond a single request or meeting?
Individual records requests and council comments matter more when they're coordinated with other people asking the same questions in the same city. MNPrivacy organizes Minnesotans working on exactly this: tracking local ALPR contracts, sharing data-request templates, comparing notes across cities, and showing up together at the meetings where these decisions actually get made instead of one resident speaking alone. Join the work at mnprivacy.org/join, and find or start a local group near you at mnprivacy.org/local.
SOURCE: [JOIN THE WORK →]
LIVE Q&A TERMINAL