Viewed from the UK, it seems America is incapable of having a reasoned discussion about anything. And the obsession with the Constitution and what the Founding Fathers might have thought defies belief. Times change, technologies change, attitudes change - and all need debating. But, I fear, we are going your way.
The 2 No / unsubscribe comments don't exactly show much ability / inclination to weigh things up.
It feels to me like cancel culture is still alive and well, at least for a vocal and irritating minority of Americans. It really repels me from politics when so many people will attack you for having any nuanced views on anything.
Thanks for the perspective on this. Powerful tools like this are always a trade off. One hopes that America can have a substantive discussion about the positives and negatives.
1. People are right to be anxious and on guard about this sort of thing. There’s no expectation of privacy for my car driving on a public road, nor do these cameras interfere with law abiding citizens’ lives (they help me). But there could be a slippery slope especially in such a hyper-partisan environment.
2. We really need to curtail qualified immunity. The incentives and culture it creates are corrosive.
You are right on point 1, I think. Of course there could be a slippery slope, without guardrails or safeguards.
I’m content that if almost all of the cops found to have abused it have been fired and/or arrested, like the Institute for Justice admits, there is a decent enough framework to prevent misuse.
On Point 2, I think there is likely to be a big QI case in SCOTUS within the next few years.
The slippery slope argument against ALPRs bothers me - I've seen it as "what if they redefine what 'wrong' is". If they redefine what "wrong" is in a world without ALPR's, then it's still a bad law, just weakly (often selectively) enforced. I would rather fight the laws themselves, and trust that when a law is universally enforced, voters have a better chance of realizing what they asked for.
to be fair, flock is not just an ALPR, but also uses video to capture all details of the car. They also offer continuous video, available for police playback, unlike other ALPRs. It also shares data into a central national pool rather than just a local department calling for the FBIs api key.
Even if ALPRs or toll transponders are not privacy-violating in a nutshell, surely the technology can become so big as to violate citizen privacy. to use your example, a cop following someone for a bit on a hunch might be ok, but following that person around all day and mapping their every move requires heavy justification even if the cops are doing the same kind of thing.
I think the ‘fingerprint’ of a car that Flock records is no more private or different to a license plate, i.e. colour/make/presence of a roof rack. It is also, from my anecdotal experience, really not very accurate.
The continuous video feature is also not available without requesting it from Flock and documenting a reason for it, it is not something an officer can simply look at default. What an officer sees day-to-day is a single frame picture.
On the last point, I think you are right - in that the Supreme Court acknowledges a cop can follow someone to find a traffic violation to stop them pre textually, but only for what might be considered a reasonable time. As of now, Flock only observes 1,700 miles of America’s 4 million miles of roads - I may change my view if that number becomes very different, but for now I don’t think anyone can suggest the level of intrusion is continuous or unavoidable.
my worry is not about individual cops abusing the systems, but the power imbalance that comes with departments having so much information. This also means I'm equally worried about the continuous video feature, and is also the worry with the national database. Flock has 120,000 cameras, with information from all of them being freely available to all 2,000 police departments.
The worry with flock, additionally, is precisely how scalable it is. Americans are worried precisely about the worry you mentioned: the number becoming very different.
This is also not to mention the drones they plan to roll out.
my point was just that it seems reductive to only focus on the ALPR technology within the article, since that I think very few Americans are worried about just that police departments are using ALPRs.
I think you are maybe misunderstanding the continuous video feature.
Flock sells CCTV video cameras, and Flock sells ALPR cameras. Continuous video is captured on the ALPR devices but only to record plates/cars as individual clips - those clips aren’t provided by default to departments only single frames, but the whole clip can be requested for a specific investigative purpose. There is no ability for a department to view the continuous video the ALPR camera captures.
If the retention period is set by Flock to be seven days, and departments must enter mandatory search reasons and case numbers, what do you think they are able to obtain from the data?
> but the whole clip can be requested for a specific investigative purpose
So that means the data is somewhere, and that's what I think most people are upset about. That our day to day movements can be recorded by a private company and then sold back to our public services. As a software engineer that has been around startups for the past 20 years, this data is the product and when a16z decides they need a higher valuation before the IPO we'll see more and more features that exploit this data that none of us had thought about. Finding stolen cars is a nice story that people can maybe get behind, but we're naive if we think that that is the end goal of this product.
The challenge is that there are only so many LEO's out there. Yes, a LEO can follow you for a long time but there is a limit. This system puts eyes everywhere. The capability is staggering.
Everything described in this article as something that flock doesn't do is an over the air software update away from being something it does do. I was eager to read an interesting counterpoint to the current moral panic but this is basically a cop saying "well it works pretty well for us".
Possible that WiP's UK perspective made this seem reasonable but in the US a traffic stop gone wrong easily and often results in the death of the person being stopped. You can in that sense maybe understand why this endorsement rings so hollow.
I'm against the death penalty for murder even after being convicted by a jury of ones peers. In the US we regularly see the death penalty for running a red light, failing to stop, speeding—on roads designed for speeding—because of an officer escalating the stop.
That takes loved ones away from their families, it harms the public trust in law enforcement, it tears the social fabric.
This is a serious editorial misstep from WiP and I would hope we hear from them about it.
“Possible that WiP's UK perspective made this seem reasonable but in the US a traffic stop gone wrong easily and often results in the death of the person being stopped” - this is the exact point made in the article and by @Sam Bowman when he shared it. Flock allows more precise, and less invasive, policing.
This may be more precise but I think it's hard to make the case that it's less invasive and an earnest, good faith reckoning with it should have been in the article, not the comments.
Cops made mistakes before Flock through incompetence/laziness/good faith, and will do so if Flock gets banned. You can find plenty of stories of people misreading license plates with their own eyes, and I guarantee a Flock camera misreads less than a human.
I don't think this response is engaging with my comment above and your replies are very fast. I think fundamentally I just expect more nuanced and considered thinking from anything promoted by WiP.
"often results in the death", saying something like this is clearly going to isolate a neutral party from your side. It is obviously not true and not helpful.
I mean, not only are fatal police shootings pretty rare when you compare it to police contact numbers, but framing them as the "death penalty" for running a red light is disingenuous. You do have examples of police officers doing their job wrong and escalating a non-violent situation into lethal violence, but the vast majority of police shootings are not the result of a simple traffic violation, they are a result of the subsequent conduct of the suspect rising to an imminent threat of death or serious bodily injury. I don't think it's fair to say that if an officer stops a car for speeding, and then the driver jumps out with a handgun to murder the officer and the officer wins the gunfight, that it's an example of an officer giving him the death penalty for speeding.
With regards to the concern of this with ALPR systems, I think that was addressed in the article. Cops screwing up is going to happen with or without flock, but flock cameras are a tool to increase the likelihood that officers are stopping vehicles actually involved in crime, as opposed to any car that looks like what the witness half remembers.
There's a tension here with your claim that internal audits will be sufficient to control abused, while also talking about how small and understaffed departments are. I work with local governments on risk management (although not law enforcement), and small agencies regularly struggle with these kind of audit and oversight functions due to lack of staffing and overall low levels of professionalism.
Which is why Flock is an advantage. The auditing is done by a well-funded tech company that is able to pinpoint abuse and use that requires further explanation. Small agencies very rarely audit use of their own software, or even CJIS systems.
Your argument is that Flock is going to audit its own customers and then report on them to some kind of higher authority? C'mon. That would be particularly unbelievable just from an economics incentive POV if the customers weren't the state security services
Accusing someone of being a secret paid propagandist is as bad-faith as you can be. It offers no recourse to the author to defend themselves and contains no substance to which anyone may reply.
It is insulting, poisonous to public debate, and smacks of an authoritarian and conspiratorial posture. Behavior like this is shameful and should be expunged from democratic debate, and that should start with you making an apology and removing or editing your comment.
No, note the words “sounds like”. And I’m sorry, but you are very naive if you think these things aren’t happening all the time. That’s the “democratic debate” issue you may want to pay more attention to.
A practice that used to be benign can assume a completely different character when technology powers it at scale. Our laws around privacy are calibrated around obstacles: cameras are expensive, storing footage is cumbersome, reviewing and analyzing it requires a human to sit down and look at it. A world of ubiquitous cameras feeding AI-analysis looks very different than the world our privacy regime was built for, and some basic assumptions about your (lack of) right to privacy in a public may have to be reconsidered. It sure looks like in the future it will be practically impossible to go anywhere without creating a police-searchable record of your location. That's just a qualitatively different power than a cop being allowed to observe license plates on passing cars, and the real argument is over whether that kind of citizen monitoring is acceptable.
The article’s Fourth Amendment analysis is far too simplistic. It assumes that because license plates are displayed in public, the constitutional inquiry is effectively over. The problem now is that the police can search through millions of records rather than watching one suspect’s license plate at a time.
The SCOTUS has not considered a case on how millions of such license plate records fare under the Fourth Amendment. Cases like United States v. Jones (GPS data aggregation for one single defendant) and the recent Geofence case are not in favor of your perspective.
There is a reason why we suppress evidence that run afoul of the Fourth and Fifth Amendments in this country. We don’t catch bad guys at the price of civil liberties. This is an issue up to the courts and the legislature. The society will eventually have to make a choice. I am not as convinced as you are.
Regardless of whether Flock has data protections in place (I firmly believe they do not.), people are not going to trust local PDs. We can argue about whether a given pct of cops are “bad” - but we have devolved into a fundamentally distrustful society.
The sudden social media fight about Flock is creepy, but I've been thinking hard about it for a year because my spouse is very anti-ALPR. To me, it's all about not letting law enforcement succumb to Baumol's cost disease. I want more policing and less prisons, and I want the chance of punishment for a crime to be high, even if the punishment itself isn't that severe (i.e fewer prisons). But none of that will be possible if law enforcement isn't allowed to become as efficient as the rest of our technology-enabled life. Like string quartets, teachers, and doctors, cops will become more expensive to hire as the alternative jobs are more productive.
I imagine you’ve had to stop driving too? It is always challenging to manage the trade off between challenging crime and believing an ALPR is tracking you.
I appreciate your perspective, but I also note that some critical thinking in response to your advocacy is warranted.
For example, you write:
> The Institute for Justice, a think tank that is critical of ALPR, studied misuse of Flock in April of this year and found 51 incidents across the United States since 2024, noting that ‘Nearly all of these officers were criminally charged and lost their jobs, either by resigning or getting fired’.
When I clicked through, I found that the Institute for Justice also believes that "The dozens of cases listed below are almost certainly an undercount. Not all police misconduct gets detected, and some cases likely get resolved quietly." They note that "only a few of the dozens of analyzed cases were initially discovered through internal investigations." This is important information. If it really was only 51 incidents, and officers were always fired, that would not be a big deal. If it's 510 incidents, 51 of which we know about, that's already a different story.
Viewed from the UK, it seems America is incapable of having a reasoned discussion about anything. And the obsession with the Constitution and what the Founding Fathers might have thought defies belief. Times change, technologies change, attitudes change - and all need debating. But, I fear, we are going your way.
The 2 No / unsubscribe comments don't exactly show much ability / inclination to weigh things up.
I think you’re right. It seems no different to the data center debate in so many ways.
It feels to me like cancel culture is still alive and well, at least for a vocal and irritating minority of Americans. It really repels me from politics when so many people will attack you for having any nuanced views on anything.
Viewed from the US, it seems the UK has mistaken 1984 for an instruction manual.
Hi Eugine - I think that's a bit of an extreme view. But you've certainly got a point - which I agree with in some aspects.
A whole lot of reason going on in the UK these days, right? OlaSnackbar.
Yes. But does that disqualify an individual from commenting?
We have a different tradition than the UK on individual rights especially in terms of speech and police powers of detention.
Weird. The UK has all the cameras, but still hasn’t managed to figure out the grooming gangs.
Thanks for the perspective on this. Powerful tools like this are always a trade off. One hopes that America can have a substantive discussion about the positives and negatives.
Thank you! People will always have different opinions but I think it’s useful to get context of what Flock can actually do
I think this article is right but…
1. People are right to be anxious and on guard about this sort of thing. There’s no expectation of privacy for my car driving on a public road, nor do these cameras interfere with law abiding citizens’ lives (they help me). But there could be a slippery slope especially in such a hyper-partisan environment.
2. We really need to curtail qualified immunity. The incentives and culture it creates are corrosive.
You are right on point 1, I think. Of course there could be a slippery slope, without guardrails or safeguards.
I’m content that if almost all of the cops found to have abused it have been fired and/or arrested, like the Institute for Justice admits, there is a decent enough framework to prevent misuse.
On Point 2, I think there is likely to be a big QI case in SCOTUS within the next few years.
The slippery slope argument against ALPRs bothers me - I've seen it as "what if they redefine what 'wrong' is". If they redefine what "wrong" is in a world without ALPR's, then it's still a bad law, just weakly (often selectively) enforced. I would rather fight the laws themselves, and trust that when a law is universally enforced, voters have a better chance of realizing what they asked for.
to be fair, flock is not just an ALPR, but also uses video to capture all details of the car. They also offer continuous video, available for police playback, unlike other ALPRs. It also shares data into a central national pool rather than just a local department calling for the FBIs api key.
Even if ALPRs or toll transponders are not privacy-violating in a nutshell, surely the technology can become so big as to violate citizen privacy. to use your example, a cop following someone for a bit on a hunch might be ok, but following that person around all day and mapping their every move requires heavy justification even if the cops are doing the same kind of thing.
I think the ‘fingerprint’ of a car that Flock records is no more private or different to a license plate, i.e. colour/make/presence of a roof rack. It is also, from my anecdotal experience, really not very accurate.
The continuous video feature is also not available without requesting it from Flock and documenting a reason for it, it is not something an officer can simply look at default. What an officer sees day-to-day is a single frame picture.
On the last point, I think you are right - in that the Supreme Court acknowledges a cop can follow someone to find a traffic violation to stop them pre textually, but only for what might be considered a reasonable time. As of now, Flock only observes 1,700 miles of America’s 4 million miles of roads - I may change my view if that number becomes very different, but for now I don’t think anyone can suggest the level of intrusion is continuous or unavoidable.
my worry is not about individual cops abusing the systems, but the power imbalance that comes with departments having so much information. This also means I'm equally worried about the continuous video feature, and is also the worry with the national database. Flock has 120,000 cameras, with information from all of them being freely available to all 2,000 police departments.
The worry with flock, additionally, is precisely how scalable it is. Americans are worried precisely about the worry you mentioned: the number becoming very different.
This is also not to mention the drones they plan to roll out.
my point was just that it seems reductive to only focus on the ALPR technology within the article, since that I think very few Americans are worried about just that police departments are using ALPRs.
I think you are maybe misunderstanding the continuous video feature.
Flock sells CCTV video cameras, and Flock sells ALPR cameras. Continuous video is captured on the ALPR devices but only to record plates/cars as individual clips - those clips aren’t provided by default to departments only single frames, but the whole clip can be requested for a specific investigative purpose. There is no ability for a department to view the continuous video the ALPR camera captures.
If the retention period is set by Flock to be seven days, and departments must enter mandatory search reasons and case numbers, what do you think they are able to obtain from the data?
> but the whole clip can be requested for a specific investigative purpose
So that means the data is somewhere, and that's what I think most people are upset about. That our day to day movements can be recorded by a private company and then sold back to our public services. As a software engineer that has been around startups for the past 20 years, this data is the product and when a16z decides they need a higher valuation before the IPO we'll see more and more features that exploit this data that none of us had thought about. Finding stolen cars is a nice story that people can maybe get behind, but we're naive if we think that that is the end goal of this product.
The challenge is that there are only so many LEO's out there. Yes, a LEO can follow you for a long time but there is a limit. This system puts eyes everywhere. The capability is staggering.
Everything described in this article as something that flock doesn't do is an over the air software update away from being something it does do. I was eager to read an interesting counterpoint to the current moral panic but this is basically a cop saying "well it works pretty well for us".
Possible that WiP's UK perspective made this seem reasonable but in the US a traffic stop gone wrong easily and often results in the death of the person being stopped. You can in that sense maybe understand why this endorsement rings so hollow.
I'm against the death penalty for murder even after being convicted by a jury of ones peers. In the US we regularly see the death penalty for running a red light, failing to stop, speeding—on roads designed for speeding—because of an officer escalating the stop.
That takes loved ones away from their families, it harms the public trust in law enforcement, it tears the social fabric.
This is a serious editorial misstep from WiP and I would hope we hear from them about it.
“Possible that WiP's UK perspective made this seem reasonable but in the US a traffic stop gone wrong easily and often results in the death of the person being stopped” - this is the exact point made in the article and by @Sam Bowman when he shared it. Flock allows more precise, and less invasive, policing.
https://www.eff.org/deeplinks/2025/10/flock-safety-and-texas-sheriff-claimed-license-plate-search-was-missing-person-it
This may be more precise but I think it's hard to make the case that it's less invasive and an earnest, good faith reckoning with it should have been in the article, not the comments.
Cops made mistakes before Flock through incompetence/laziness/good faith, and will do so if Flock gets banned. You can find plenty of stories of people misreading license plates with their own eyes, and I guarantee a Flock camera misreads less than a human.
I don't think this response is engaging with my comment above and your replies are very fast. I think fundamentally I just expect more nuanced and considered thinking from anything promoted by WiP.
"often results in the death", saying something like this is clearly going to isolate a neutral party from your side. It is obviously not true and not helpful.
sorry, being loose with language. Often in this case meaning multiple times higher per capita compared to other developed countries.
I mean, not only are fatal police shootings pretty rare when you compare it to police contact numbers, but framing them as the "death penalty" for running a red light is disingenuous. You do have examples of police officers doing their job wrong and escalating a non-violent situation into lethal violence, but the vast majority of police shootings are not the result of a simple traffic violation, they are a result of the subsequent conduct of the suspect rising to an imminent threat of death or serious bodily injury. I don't think it's fair to say that if an officer stops a car for speeding, and then the driver jumps out with a handgun to murder the officer and the officer wins the gunfight, that it's an example of an officer giving him the death penalty for speeding.
With regards to the concern of this with ALPR systems, I think that was addressed in the article. Cops screwing up is going to happen with or without flock, but flock cameras are a tool to increase the likelihood that officers are stopping vehicles actually involved in crime, as opposed to any car that looks like what the witness half remembers.
On the plus side, all these places removing it at different times will make it much easier for some economists to prove how well they work.
There's a tension here with your claim that internal audits will be sufficient to control abused, while also talking about how small and understaffed departments are. I work with local governments on risk management (although not law enforcement), and small agencies regularly struggle with these kind of audit and oversight functions due to lack of staffing and overall low levels of professionalism.
Which is why Flock is an advantage. The auditing is done by a well-funded tech company that is able to pinpoint abuse and use that requires further explanation. Small agencies very rarely audit use of their own software, or even CJIS systems.
And Flock has every incentive to say that its being used appropriately. Come on now. All the while you are saying it is super cheap.
Your argument is that Flock is going to audit its own customers and then report on them to some kind of higher authority? C'mon. That would be particularly unbelievable just from an economics incentive POV if the customers weren't the state security services
Yes, the article sounds like The Works in Progess took money from Flock (is this an advertorial?)
No such thing has taken place, nor has any money been paid by Flock directly or indirectly to me or the publication.
Sure officer. We believe you.
Accusing someone of being a secret paid propagandist is as bad-faith as you can be. It offers no recourse to the author to defend themselves and contains no substance to which anyone may reply.
It is insulting, poisonous to public debate, and smacks of an authoritarian and conspiratorial posture. Behavior like this is shameful and should be expunged from democratic debate, and that should start with you making an apology and removing or editing your comment.
No, note the words “sounds like”. And I’m sorry, but you are very naive if you think these things aren’t happening all the time. That’s the “democratic debate” issue you may want to pay more attention to.
Yeah, weasel words are for weasels. Accusations should be made with evidence.
Conflict of interest statements are a standard ethical principle. It’s not a new concept. It’s also not hard at all.
A practice that used to be benign can assume a completely different character when technology powers it at scale. Our laws around privacy are calibrated around obstacles: cameras are expensive, storing footage is cumbersome, reviewing and analyzing it requires a human to sit down and look at it. A world of ubiquitous cameras feeding AI-analysis looks very different than the world our privacy regime was built for, and some basic assumptions about your (lack of) right to privacy in a public may have to be reconsidered. It sure looks like in the future it will be practically impossible to go anywhere without creating a police-searchable record of your location. That's just a qualitatively different power than a cop being allowed to observe license plates on passing cars, and the real argument is over whether that kind of citizen monitoring is acceptable.
Zvi Moshovitz calls this "Levels of Friction". https://thezvi.substack.com/p/levels-of-friction
The article’s Fourth Amendment analysis is far too simplistic. It assumes that because license plates are displayed in public, the constitutional inquiry is effectively over. The problem now is that the police can search through millions of records rather than watching one suspect’s license plate at a time.
The SCOTUS has not considered a case on how millions of such license plate records fare under the Fourth Amendment. Cases like United States v. Jones (GPS data aggregation for one single defendant) and the recent Geofence case are not in favor of your perspective.
There is a reason why we suppress evidence that run afoul of the Fourth and Fifth Amendments in this country. We don’t catch bad guys at the price of civil liberties. This is an issue up to the courts and the legislature. The society will eventually have to make a choice. I am not as convinced as you are.
Regardless of whether Flock has data protections in place (I firmly believe they do not.), people are not going to trust local PDs. We can argue about whether a given pct of cops are “bad” - but we have devolved into a fundamentally distrustful society.
The sudden social media fight about Flock is creepy, but I've been thinking hard about it for a year because my spouse is very anti-ALPR. To me, it's all about not letting law enforcement succumb to Baumol's cost disease. I want more policing and less prisons, and I want the chance of punishment for a crime to be high, even if the punishment itself isn't that severe (i.e fewer prisons). But none of that will be possible if law enforcement isn't allowed to become as efficient as the rest of our technology-enabled life. Like string quartets, teachers, and doctors, cops will become more expensive to hire as the alternative jobs are more productive.
Unsubbing. This is advocating a surveillance state.
i am unsubscribing because of this article.
I imagine you’ve had to stop driving too? It is always challenging to manage the trade off between challenging crime and believing an ALPR is tracking you.
Why?
Automated sensor networks always create weird emergent surveillance trade-offs. The real question is how state retention and query logs are audited over time. Covered information theory and system boundaries in this piece: https://gzambrano.substack.com/p/the-physics-of-what-a-model-decides-to-forget-information-approximation-and-the-limits-of-simulation
I appreciate your perspective, but I also note that some critical thinking in response to your advocacy is warranted.
For example, you write:
> The Institute for Justice, a think tank that is critical of ALPR, studied misuse of Flock in April of this year and found 51 incidents across the United States since 2024, noting that ‘Nearly all of these officers were criminally charged and lost their jobs, either by resigning or getting fired’.
When I clicked through, I found that the Institute for Justice also believes that "The dozens of cases listed below are almost certainly an undercount. Not all police misconduct gets detected, and some cases likely get resolved quietly." They note that "only a few of the dozens of analyzed cases were initially discovered through internal investigations." This is important information. If it really was only 51 incidents, and officers were always fired, that would not be a big deal. If it's 510 incidents, 51 of which we know about, that's already a different story.
It's like saying "we issued 1,000 speeding tickets last year, so that's how many speeders there were"!