If you want to save yourself six minutes' of video, this is about https://alpr.watch/ and their new feature that can alert you by email if your local municipal officials are going to be discussing Flock in upcoming meetings, based on published meeting agendas.
The video also links you to a wiki with some nice counter-arguments to the standard pro-Flock arguments: https://consumerrights.wiki/w/Common_Questions,_Arguments,_%...
I went ahead and signed up; I live in a pretty dense part of the US, we'll see how many alerts I get in the next year.
Even in my local community, which is likely among the 10 most progressive in the country (we're either the first or second most progressive in Chicagoland, itself one of the most reliably blue major metros in the country), support for Flock was pretty evenly divided.
I think it's good to engage this way, but I have a lot of thoughts on things to do that are more effective than giving public comment, and a caution that if you have strong opinions about ALPRs and you choose to pay attention to this issue you're going to be confronted with a lot of opinions that may surprise/discomfit you.
I strongly suspect that opposition to surveillance technology doesn't manifest along a simple progressive-conservative divide.
In my experience, a sizable chunk of people who are anti-surveilance are pretty staunchly rightwing.
This is bad news in that it means that there isn't a pre-formed anti surveillance coalition, but good news in every other way imo.
At the last public meeting that someone from our household attended, people screamed and loudly booed during people's time to address the local elected officials.
So, I'm aware that the people I live next to are tremendously rude dipshits who hold awful opinions.
https://deflock.me/ deserves a mention for its crowdsourcing of ALPR camera data on https://openstreetmap.org and on the site. Recording even one camera may be the only notice a resident has that Flock and ALPRs are operating in their municipality.
Well, I didn't know there Flock cameras in use near me, but apparently I'm nearly surrounded and would have to take a weird route to avoid them. Some are marked as being operated by the local PD, and others are "Unknown". Thanks for the link
Interesting. All the Flock cameras around me are stationed around the entrances to Lowe's parking lots.
Lowe's and Home Depot both seem to be hubs for their cameras. I only know of one in my rural area and it's at the Lowe's entrance.
Home Depot and Lowe's Share Data From Hundreds of AI Cameras [Flock] With Cops - https://news.ycombinator.com/item?id=44819750 - August 2025
All the Flock cameras around me are stationed around the entrances to Lowe's parking lots.
Most of the ones in my neighborhood are pointed at parks, playgrounds, and the big transit center. Which makes no sense to me since there's a ton of government buildings around that you'd think would be under Flock surveillance for "safety."
All of the ones I've noticed have been pointed directly towards streets for mostly license recognition but it's notable that they record whatever objects a typical real world AI image model could. In my area, we have Flock, Shotspotter, Stingray devices, free Ring camera programs from law enforcement departments.
Our Lowe's have the mobile parking lot camera/light units, I wasn't aware if these were Flock but either wouldn't be surprised if they were, had access or plans to buy in.
Same author talked about adversarial license plates that trick these cameras with a sequence of black blocks, discussed here in original form [1]. He is interested in breaking both the plate detection (ideal) and character recognition (good). The examples are pretty cool looking.
In most countries, this is prohibited by law. While it might be interesting from a technical perspective, it does not help in practice.
Yep, and the overwhelming majority of people using them are not principled cypherpunks, but parking fee dodgers and habitual dangerous drivers.
Instead of a sticker like in the video make a stencil and spray diluted mud through it. Plausible deniability!
Are you also going to spray your car with mud too? Going to have a hard time explaining a spotless car that only has mud on the license plate.
Many police cars now have ghost graphics.
https://gdigraphics.com/police-car-ghost-graphics/
There were laws in many places where you could fight a traffic ticket because you couldn't plainly recognize a police vehicle, especially when a taillight or headlight is out, but now we pay for graphics to make them more invisible. "If you have nothing to hide, you have nothing to worry about." I like the plausible deniability angle, myself
My car is 'self-spraying' so much I'd like it to be less so. Country life I guess.
Not a problem, the TPMS will give you away.
Get it while it's hot, cuz it's already illegal in some states, and will be in more soon!
You will be tracked and you will be happy about it.
Flock data retention is defaulted to 30 days, but can vary up to a year or longer depending on the terms of the municipality contract.
Is this retention period configured in Flock’s data lake by camera? Or by entity or agency the camera is assigned to?
I wonder what would happen to these if you put a bunch of TV screens showing random faces at various camera locations. Essentially creating 10s of thousands of face scans per minute at each location until their database fills up and their facial recognition runs out of CPU cycles. Also maybe throw in some randomized license plate numbers and a TPMS transmitter to make it even worse. Nothing illegal, just putting some noise out there.
Flock doesn't do facial recognition.
Even if you’re technically correct, that doesn’t mean they never will, or that downstream consumers of their data never will. They’re easy enough to hack and I’m sure it would be trivial for the NSA to siphon off the raw feed in real time if they wanted to and send it to a datacenter in Utah.
I was answering the parent commenter's question about what would happen if you confronted a Flock camera with a lot of faces. Answer: nothing.
Oh, yeah, looks like they only do general description on pedestrians. Stuff like orange sweater and black pants.
yet
I am always on the fence with flock. I can absolutely see how it goes wrong and wish there was more oversight into the ability to track people. On the other hand I see these as a very effective way to assist local law enforcement who across the country are struggling with budgets.
It feels like there’s a big vacuum of a federal level privacy judgement waiting to be filled. SCOTUS abhors a vacuum, though you could argue it’s already been filled in that the past two decades of rulings on cell phone records (phone=you) are directly applicable to photos of your car (car=you).
Maybe the argument against is that ALPR can’t constantly track you like cell towers can?
I can't see the link due to corporate restricted mode for youtube, is this regarding the same Flock that is a YC company?
Quiet red areas rolling over to Flock are what’s going to cause us all to lose in the end.
Another quiet little village in rural New York just signed on for 11 cameras, and it sounds like the county itself (2800 square miles(!)) is also playing around with them. The locals won’t raise the hard hitting questions - they’ll just roll over with the bullshit answers from Flock reps.
https://northcountrynow.com/stories/village-of-massena-enter...
Forget "quiet red areas". Every suburb surrounding mine, in Chicagoland(!), has expanded their use of ALPR cameras. People want them, not just in red areas.
Blue areas have often failed to do a good job suppressing crime (e.g. California), so people seek alternatives.
I have my complaints about this, but "blue areas" over the last several decades have meant "all major metros", and some forms of crime are just endemic to density, and in some sense all crime simply tracks population.
But we don't have to agree on that; people everywhere care about crime, and the promises ALPR vendors make, while arguable, are not ludicrous.
In my area, for example, there are counties where if a thief can escape there, they will not be prosecuted. So I support Flock since it gives a higher chance of interception in my county, which will prosecute them.
What we found is that the cameras mostly had us enforcing failure-to-appear warrants for neighboring municipalities, which was not an in-kind contribution we wanted to make to those munis, and so we didn't get much value from the cameras, which is a big part of why we shut them off.
> Quiet red areas rolling over to Flock
What is funding all those Flock reps jetting around BFE to dazzle and kickback the boomer city managers and county commissioners of deep red littleville America? Is it the 2 cameras in Big Rapids MI or the 2425[1] cameras in Detroit metro?
The "roll over" that mattered has already been secured.
Do flock reps even need to fly out? They have massive contracts with the Walmarts of the world and the underlying commercial property owners. You don’t need to have a rep when it’s already in your area.
The anti ALPR narrative is not based in reality. I for one support using tech to automatically flag when a stolen car is spotted. With the sky high cost of car insurance in CA, which disproportionately impacts low income drivers, you would think liberal legislators would be in favor of reducing one of the largest reasons insurance is expensive. Restricting tech used by police just means more LE time spent on easily automatable tasks, and forces LE to use their own judgement (which many would argue has bias). The ACLU and EFF are so discredited on tech issues. They simply support criminals. The ACLU is fighting DUI laws in CA right now for instance. SF is a hell-hole because of these crime loving activist groups.
Thank you, Flock!
> SF is a hell-hole because of these crime loving activist groups.
the cost per arrest in austin while these cameras were up was over $7000/arrest.
From an anti-crime perspective, spending $7000 on police salaries & having the police patrol high crime areas is a much more effective use of funds. It also doesn't violate the 4th amendment.
I can't wait for the day when Flock's "proactive AI" flags the way you are driving or your vehicle movements as suspicious and alerts LE to just ... "check in on you".
Or when they enable the mics in their devices to just start recording your conversations with your friend in a public place and does the same. "AI didn't like what you were talking about, so alerted the local PD".
> alerts LE to just ... "check in on you".
This is currently an epidemic. Drivers are targeted for “random” checks by police for a number of non-falsifiable factors (e.g. the evergreen “your license plate light was out…huh, looks fine now”) that overwhelmingly correlate with driver income and race.
That’s not whataboutism; I am genuinely not sure if ALPR/automated policing systems stand to make that situation worse or better. Are Flock and friends likely to be abused in the same way that human police traffic stop reasons are?
I have every reason to believe so.
Flock's founders belief is that he wants to eliminate all crime (literally) with Flock.
So in his eyes, false positives are inherently acceptable, and preferable to false negatives.
And I feel that (actually, I know that, though I wasn't in Sales, but I did work at Flock) one of their selling points to agency is almost a "whitewashing" of such practices. "Oh, our PD wasn't targeting anyone, we were just acting on the recommendations of the Flock surveillance system".
I am not anti ALPR. They do have merits on flagging things like stolen vehicles and human trafficking. What I do dislike is a private company owning up the entire market share and has little to no obligation to behave.
Why can't cities hire good software developers to create custom solutions that are safe and secure rather than paying a startup thousands of dollars in taxpayer money. Austin City Council spent 1.2 MILLION dollars on just a handful of cameras. Texas already ruled that red-light cameras cannot issue tickets or citations so why are we allowing cameras to creep into the same space. It's just another tax on people
>The ACLU is fighting DUI laws in CA right now for instance. SF is a hell-hole because of these crime loving activist groups.
Not sure what you are referring to here but if you do find the news stories be sure to post the article and I will read the cases on PACER to see what its all about
We don’t need flock to track stolen or flagged vehicles, there is already a national patchwork of cameras (typically on intersections) that do this, bullet type, dome type, and the very old IR ALPR type. Been this way a long time.
All flock gets you is more of that except also every petty theft gets run and then they harass an old lady. It almost certainly costs more, too.
https://www.cbsnews.com/amp/colorado/news/flock-cameras-lead...
I would be surprised it costs more. I am sure there are false positives but flock is generally pretty effective for local pd. I for one would rather have them to save police resources. This can also help in not pursuing folks that run. You get the license plate and you will see it soon enough.
There are no reliable studies. There have been multiple demonstrated examples that user data has been shared, and systems have been hacked.
We have plenty of evidence of harm and no good evidence for effectiveness. We don't need to "save police resources" we need more well-trained capable police officers who are doing good police work AND good community work
"Austin spent $1.2 million on Flock cameras. They scanned 113 million license plates and got 165 arrests. That's $7,300 per arrest"
I had to go back and look. The quote you have does not seem correct. That 165 is coming from an audit for a trial of both flock and axon patrol car cameras. Only 40 flock cameras. Flock has pretty transparent pricing it’s about $3k per camera for cities. So let’s call it $150k for that test. They installed patrol car cameras in 500 cars. On volume alone the cost is with the cars not flock.
Again I don’t know where that 1.2 million number is coming from. That should get you over 300 cameras deployed in the first year.
7300 per arrest sounds cheap so I am not sure if you are for or against.
I agree there needs to be better safe guards. I still believe it’s worth figuring out a balanced path forward, I like having cameras track public streets.
>7300 per arrest sounds cheap so I am not sure if you are for or against.
think about what a police officer's salary is. think about underpoliced areas.
how many arrests would a trained, qualified police officer make in a high crime area of a major city in the timeframe it takes for them to earn $7300?
this is a bad deal.
Maybe? But the bigger problem is this number of 7300 is being made up in thin air. It’s bad math because it’s conflating too many things in the underlying audit.
>Not sure what you are referring to here but if you do find the news stories be sure to post the article and I will read the cases on PACER to see what its all about
https://calmatters.org/investigation/2025/12/california-road...
Has this tech lowered car theft rates and insurance rates?
Nope.
Should mention, stealing nearly any car is extremely easily, and quite fast, too.
The reason more cars aren’t stolen is because they are registered to an owner, and the resale value of a stolen cars is on the order of a couple hundred bucks USD for a brand new vehicle. It simply isn’t worth the time.
Essentially nobody is stealing cars except for a few chop shop operations who concentrate in specific areas.
I am personally not a fan of the taste of leather. But to each their own.
Alternate parse: "we put flock(2) under surveillance: go(1) makes them behave differently"