What EFF is asking for
The demand is narrow and old. It is not a ban, and it is not aimed at the cameras reading plates in real time.
The ask, in one line
A warrant, signed by a judge
EFF's phrasing puts a person in the loop deliberately. An audit log catches a bad search after the fact. A judge reviewing the request catches it before the search happens at all.
The word doing the work is "historic". Reading a plate as a car passes is one thing; querying months of stored sightings to reconstruct where that car has been is another, and EFF treats the second as the point where judicial review has to attach.
"To be clear, our position has long been that police, at a minimum, need to get a warrant, signed by a judge, in order to search for historic ALPR data regarding specific vehicles. For us, it's common sense: if police want to dip into historic ALPR data like they were going back in time to retroactively follow your comings and goings, they need a warrant."(EFF's position)／"He added that he has come to agree with groups like the American Civil Liberties Union and the Electronic Frontier Foundation that police should need an active case number to search Flock's data."(on Flock's CEO) — EFF Deeplinks
Why a stored-data search is retroactive tracking
A live camera answers where a car is now. A database of stored sightings answers where your car was on any day it passed a camera, which is a different capability wearing the same name.
That difference is why retention length matters so much in this debate. Shortening the window does not change the nature of the search; it changes how far back the search can reach.
What Flock Safety changed
The package is real, and EFF says so. Retention got shorter, and the audit feature is described as strengthened. The split is over what that adds up to.
The reform package
Data retention: 30 days down to 7
This is the change EFF calls the most consequential, and the reason is arithmetic, not principle. A shorter default window means fewer historic sightings sitting in the system for any future query to find.
The qualifier in the wording is "optional". A default is a starting position that a customer can move, so the protection holds for departments that leave it alone and does nothing for those that do not.
"The most consequential is Flock's default setting of an optional 7-day retention period for ALPR data, down from its original default optional 30-day retention period. This means if police want to retain data beyond the duration of their retention setting, they need to access "Evidence Mode," i.e., when the desired data is associated with an active investigation and not just a fishing expedition." — EFF Deeplinks
Offense filtering and audit lockouts
Two further changes address who else can reach a city's data and what happens when a request looks wrong. Offense filtering lets a city expose its data for murder or robbery investigations while shutting out, say, immigration ones. The second is what EFF calls a supposed strengthening of the audit feature, with automatic lockouts for officers filing suspicious requests.
EFF does not call the auditing useless. The complaint is that the tooling fixes a problem the product itself created.
"Flock Safety, the embattled vendor of mass surveillance technology, has rolled out a handful of new reforms intended to appease the justified nationwide anger that has seen scores of towns cancel or suspend their contracts with the company for automated license plate readers (ALPRs)."(opening)／"The first is offense filtering so that cities can enable other departments to access their ALPR data only if they are investigating certain crimes, e.g., murder or robbery but not immigration-related investigations."(the reforms)／"The second is supposedly beefing up their audit feature and proactively locking out officers who file suspicious requests for data."(the reforms)／"The major problem here is the fact that Flock's enhanced audit and transparency tools help to address a problem that Flock itself has created—an abusable mass surveillance system that tracks all cars all the time."(EFF's objection) — EFF Deeplinks
Why a vendor cannot settle this
The strongest part of EFF's argument is not about any individual setting. It is about who holds the pen.
Two ways to set the same rule
Nothing stops a reversal
A protection a company grants is a protection a company can withdraw. EFF puts the question directly: if departments start defecting to a competitor with looser terms, what holds these settings in place.
The answer it gives is one word, and the argument does not depend on predicting bad faith. A vendor responding to its market is behaving normally; the problem is that a civil liberties floor was left resting on that response.
"Fundamentally, these reforms leave us wondering: what is stopping Flock from reversing course on them if their law enforcement customers respond by defecting to another ALPR vendor? Nothing."／"It shouldn't be up to Flock or any other ALPR vendor to decide how long police can collect and retain data on millions, if not hundreds of millions, of innocent people." — EFF Deeplinks
The fix has to be legislative
EFF's closing position moves the whole question out of the product and into law: restrict police use of surveillance technology by statute, because the business model is what generates the exposure.
That reframing is what makes the reform package read as an admission, not a resolution. A company conceding that its defaults were wrong has conceded the premise, and the remaining disagreement is only about who gets to set them next time.
"We need lawmakers to step up and pass laws that restrict police's use of surveillance technology. After all, the surveillance business model is the problem, and a few company-imposed slapdash reforms aren't going to change that." — EFF Deeplinks
For the same argument playing out in statute instead of vendor policy, see Colorado AI Act and EU AI Act: what changes in August 2026.
Advocacy posts and vendor announcements are easier to read against each other in plain text.
