How Federal Agencies Access Flock License Plate Data Without Contracts or Warrants
The shadow surveillance pipeline: how federal agencies get Flock access
Flock Safety operates more than 500,000 automatic license plate reader cameras across North America. Those cameras capture over 100 million plate reads every day, logging the time, location, and image of every vehicle that passes. Federal agencies are using that infrastructure — and they built a system specifically designed to avoid leaving fingerprints.
The mechanism is indirect by design. ICE and CBP do not hold formal contracts with Flock Safety. They do not need to. Local police departments already have data-sharing agreements with Flock, and those same departments have memoranda of understanding with federal partners. The federal agency queries the local department, the local department queries Flock, and the data moves upstream. No federal procurement process. No public contract. No transparency requirement triggered.
This arrangement was not an oversight. It was engineered.
Why federal agencies don't need warrants for ALPR searches
There is no federal law governing ALPR data collection, retention, or access. Zero. States have tried to fill that gap with varying results: a handful require warrants before law enforcement can query ALPR databases, most do not, and none can prevent federal agencies from accessing data through informal channels that sidestep state-level protections entirely.
The Fourth Amendment question is genuinely unsettled. Courts have not clearly established whether querying a searchable database of vehicle location records constitutes a "search" requiring judicial approval. The government's position has consistently been that ALPR data is collected in public spaces, that driving on a public road carries no reasonable expectation of privacy, and that database queries are therefore not searches at all. That argument holds until you consider what the database actually contains: a retroactive, searchable record of everywhere a vehicle has been, potentially for five or more years.
The EFF has been pushing courts to treat ALPR database queries the way the Supreme Court treated cell-site location information in Carpenter v. United States (2018) — as a form of comprehensive location surveillance requiring a warrant regardless of where the underlying data was collected. That argument has not yet produced a binding federal ruling on ALPR specifically.
When local police hold the access and share it with federal partners, courts have historically treated that transfer as a third-party data arrangement rather than a direct government search. That framing lets federal agencies conduct mass surveillance while pointing to the local department as the party with legal authority over the data.
Flock Safety as a de facto government contractor
Flock Safety is not classified as a federal contractor. That classification matters because federal contractors face procurement rules, oversight requirements, and public disclosure obligations that Flock currently avoids entirely. The company's business model runs on local law enforcement partnerships, which create the data-sharing relationships that federal agencies then exploit.
Flock retains ALPR data for five or more years by default. No legal requirement compels that retention period, and no legal requirement compels deletion either. What that creates is a searchable archive of vehicle movements that federal agencies can query retroactively, building timelines of where a vehicle has been long before any investigation formally began.
A company holding a formal federal contract would face Freedom of Information Act exposure, Inspector General oversight, and congressional scrutiny. Flock, operating through local intermediaries, faces none of that at the federal level. Federal agencies can say they don't contract with Flock. Flock can say it only serves local law enforcement. Both statements are technically true. Neither one is the whole picture.
How ICE and CBP use this data for immigration enforcement
ALPR access changes what immigration enforcement looks like operationally. Instead of building a case through traditional investigative methods requiring some showing of probable cause, agents can query vehicle records associated with a known address, a workplace, or a travel corridor. The database returns a history of vehicle movements. That history informs raid planning, identifies patterns of travel, and flags vehicles associated with specific locations.
Probable cause of criminal activity is not required to initiate these queries. Immigration status alone is sufficient justification under the current framework. That distinction matters operationally: it means the entire ALPR infrastructure can be directed at immigrant communities based on civil immigration violations, converting a commercial surveillance network into a deportation planning tool.
The scale is impossible to audit. The number of federal ALPR queries conducted annually is unknown because no transparency requirement exists that would force disclosure. FOIA requests have revealed that thousands of local police departments share ALPR access with federal partners through informal agreements, but the scope of what federal agencies actually pull from those systems remains opaque.
The legal vacuum and why existing protections fail
State-level ALPR protections were designed to constrain local law enforcement. They were not designed for, and do not cover, federal agencies accessing data through informal channels. A state that requires a warrant before its own police can query ALPR databases cannot enforce that requirement against ICE accessing the same data through a federal-local MOU.
The privatization of the underlying infrastructure creates a second gap. Flock owns and operates the cameras. The data lives on Flock's servers. Federal agencies access private data through private intermediaries, which means the statutory constraints that would apply to government-operated surveillance systems do not apply here. The government is a customer of a customer, legally speaking, and that distance insulates the arrangement from direct constitutional challenge.
The surveillance capability is real and operating at scale. The legal framework for challenging it is built on doctrines developed for different technologies, applied to a system that falls between every existing category of oversight.
What the EFF and privacy advocates are doing
The EFF filed an amicus brief arguing that law enforcement must obtain a warrant before querying ALPR databases, framing the database search as distinct from a patrol officer incidentally observing a license plate. The argument draws on Carpenter: comprehensive location records, even if assembled from public observations, constitute a qualitatively different kind of surveillance that the Fourth Amendment should reach.
Immigrant rights organizations are building parallel evidentiary records documenting how ALPR data feeds directly into enforcement operations, with the goal of establishing standing for Fourth Amendment and due process challenges. Hard to do when the surveillance is by definition undisclosed.
Legislative advocacy is focused on federal ALPR legislation that would mandate warrant requirements for database queries, cap data retention periods, and require agencies to publish access logs. That legislation does not exist yet, and given the current political climate around immigration enforcement, it is not close.
Why warrant requirements are the minimum necessary fix
A federal warrant requirement for ALPR database queries would not end government access to private data through license plate cameras. It would require agencies to demonstrate probable cause to a neutral magistrate before accessing location history. That is a low bar by design, but it is a bar, and it would prevent the current practice of querying databases speculatively or based solely on immigration status.
Transparency mandates matter separately from warrant requirements. Requiring federal agencies to log and disclose the volume and purpose of ALPR queries would expose the actual scale of this surveillance to public scrutiny. Right now that scale is unknown. Unknown surveillance is unchallengeable surveillance.
Data retention limits would eliminate the retroactive tracking capability that makes ALPR databases particularly powerful. A system that retains data for 30 days is useful for active investigations. A system that retains data for five years is useful for building retrospective profiles of anyone whose vehicle was ever captured by a camera — which, at 100 million reads per day, is most of the country.
Without those reforms, the pipeline between Flock Safety and federal enforcement agencies will keep running through local police departments as silent intermediaries: invisible to procurement oversight, immune to state privacy law, and accountable to no one. That is not a gap in the system. That is the system working as designed. Change it or accept what it produces.