ACLU Battles Connecticut Law Enforcement Agencies Over Retention Of Licence Plate Reader Data
from the grab-it-all-and-hold-it-tight dept
Automatic license plate readers (ALPRs) are pretty much standard operating equipment for law enforcement agencies across the nation. Even if the ill-timed attempt to create a national license plate database was recently killed by DHS/ICE officials, this removal has done nothing to deter the collection efforts of multiple ALPR manufacturers to harvest tons of plate/location data and provide warrantless access to law enforcement.
Standard operating procedure also means rolling these plate readers out without asking for public comment or providing any meaningful privacy protections. Most ALPRs will hold onto ALL data for five years, unless forced to trim that down by legislators responding belatedly to public outcry. Some will hold onto the data for shorter periods of time, but nearly every ALPR manufacturer makes little to no effort to discard non-hit data.
The ACLU is currently fighting Connecticut legislators and police departments over proposed data policies relating to the states ALPRs.
The length of time law enforcement agencies could retain license plate information was the center of arguments Tuesday before the legislative Public Safety and Security Committee.The police want five years, a ridiculous amount of time, especially considering they want to keep it all for that long -- even non-hit data. As David McGuire of the ACLU points out, this extended retention period of non-hit data lends itself to abuse.
Local police chiefs want to hang on to the information for five years or more, because they say it's an important crime-fighting tool.
But opponents, including the American Civil Liberties Union of Connecticut, called the collection -- known as automated license plate recognition -- a threat to constitutional freedoms and asked the committee to require a much shorter retention period of days instead of years.
"The trouble arises when license plate scan data is collected, pooled and archived for months or years, storing a detailed and vivid picture of the movements of drivers who are not even suspected of doing anything wrong.Why would a law enforcement agency need 5 years of non-hit data? Connecticut police chiefs call the ALPRs an "important law enforcement tool" but what does plate data unrelated to investigations have to do with "law enforcement?" The ACLU suggests a retention period of two weeks for data not related to open investigations, but law enforcement officials want it all -- for a half-decade.
Kept for years in databases, plate data can create a method for "retroactive surveillance of innocent people without a warrant, without probable cause and without any form of judicial oversight," he said.
Without a doubt, ALPRs are a useful tool for law enforcement agencies, but one has to wonder why they're fighting so hard to keep the unrelated data. It serves almost no purpose and only encourages fishing expeditions and other abuses. No law enforcement rep interviewed for these stories offered a single reason why the ACLU's suggestion wouldn't work.
The only conceivable reason would be the investigation of cold cases or other incidents that happened years ago where location data might be useful. If that's the rationale, the rule should still hold. The non-hit data could be retained past the two-week point but should only be accessible with a warrant -- something narrowly crafted to search for only relevant plate numbers.
As the ACLU has discovered, even a small state like Connecticut has already amassed millions of license plate/location records. Six million records have been collected by ten towns, with Newington, CT alone compiling over 600,000 scans despite only having a population of 30,000. And for what? Newington police chief Richard Mulhall points to a little under 900 arrests (with 839 of those being "motor-vehicle related" -- only 28 were deemed "criminal arrests") as justification of the massive amount of scanning.
There's an obsession with collecting data -- a majority of it useless and irrelevant -- pervading every law enforcement, investigative and intelligence agency in the nation. Groups like the ACLU are pushing back using FOIA requests and legislation, but it's an uphill battle against this ingrained mentality. Technology continues to outpace the law and as long as this remains a fact, the privacy of millions of Americans remains threatened.
Thank you for reading this Techdirt post. With so many things competing for everyone’s attention these days, we really appreciate you giving us your time. We work hard every day to put quality content out there for our community.
Techdirt is one of the few remaining truly independent media outlets. We do not have a giant corporation behind us, and we rely heavily on our community to support us, in an age when advertisers are increasingly uninterested in sponsoring small, independent sites — especially a site like ours that is unwilling to pull punches in its reporting and analysis.
While other websites have resorted to paywalls, registration requirements, and increasingly annoying/intrusive advertising, we have always kept Techdirt open and available to anyone. But in order to continue doing so, we need your support. We offer a variety of ways for our readers to support us, from direct donations to special subscriptions and cool merchandise — and every little bit helps. Thank you.
–The Techdirt Team
Filed Under: connecticut, data, license plate readers
Companies: aclu
Reader Comments
Subscribe: RSS
View by: Time | Thread
[ link to this | view in thread ]
Anyway, considering all the VERY recent historical examples of why this is a bad idea I'd say Americans need to review what's being taught in their schools regarding History. It seems the generation currently in power fails at it.
[ link to this | view in thread ]
[ link to this | view in thread ]
Re:
Getting a warrant takes effort, it means going before a judge and presenting evidence as to why they should be allowed to gather data on a suspect, and provides limits as to just what they can gather, all of which creates a paper trail of their actions. Using tech like this however, with enough data points you can track someone quite effectively, all without that pesky 'warrant' and paper-trail stuff getting in the way.
[ link to this | view in thread ]
Re:
I mean, when you loan a tool or any other object to someone else, obviously it's your responsibility to keep proper documentation as to just what is being loaned, and who it's being loaned to, keeping the documentation for at least five years just in case it might be important at some point, anything less would be horribly irresponsible!
/s
[ link to this | view in thread ]
Re: Re:
[ link to this | view in thread ]
[ link to this | view in thread ]
Re: Re:
[ link to this | view in thread ]
Re:
Works part time job to pay for college? Why can't you do the same?
Could afford a house right out of high school with unskilled labor positions? Why kids are just lazy today!
Baby boomers never even fathom how much better they had it when they were young, where labor was valued and just having a job was considered good enough.
If they got stuck working for peanuts with no obvious path to a better life except getting saddled with $50+k in debt maybe they would change their tune...
As for the data retention, they want it to essentially track drug dealers or potential child abductors.
The former they can find a known car of a dealer, then go back and track where he tends to go in and out of. Once they establish a pattern they could start surveillance of more specific locations.
For the later they could get a partial hit off a plate or even a whole hit and try to narrow down where the suspected criminal could be. OR alternatively know when a child was taken and pool a list of plates from the time it happened.
For the former you could argue it would be helpful, but ultimately it is a shortcut to field work that needs to be done anyway. For the later there is NO reason to hold data that long(IE longer than say 30 days) other than sheer incompetence of taking the time out of your day to search through the system.
[ link to this | view in thread ]
Re: Re: Re:
Fortunately, the current state of face recognition is still pretty poor. The inaccuracy of it renders it largely useless except for certain specific applications.
[ link to this | view in thread ]
Re: Re: Re: Re:
[ link to this | view in thread ]
Re: Re: Re:
[ link to this | view in thread ]
Re: Re: Re: Re: Re:
"It might not be too hard, even with present technology, to verify if a car registered to John Q. Smith is being driven by someone resembling John Q. Smith or someone else who uses the same address."
Yes, this would be one of the specific applications. But the best that face recognition can do is to present a set of people that a photo might be of -- a human is still required to make the ID more specific.
And remember, humans do face recognition better than any machine that is in the realm of known possibilities right now -- and humans don't actually do it as well as we like to think. That we humans do voice recognition much more accurately than face recognition indicates that face recognition might, in fact, be the more difficult of the two problems.
[ link to this | view in thread ]
The argument against long term retention should be focused on the fact that this is yet another method to automatically track the movements of nearly all citizens. Currently, the majority of people will have their cars recorded, either not at all or only a few times a year. That is not so worrisome. However, these use of these systems is rapidly proliferating and when your car is recorded everyday that data, whether in the hands of government or a private company is rather worrisome.
[ link to this | view in thread ]
More Problems
I would imagine in a particularly nasty divorce, th polie would be hard-pressed to say to a judge "it's not the spouse's business to look at the travel history of their spouse or possible lovrs."
Ditto in, say, a lawsuit over wrongful dismissal - I sure hope the police clocks are fully calibrated (and configured for the ight time zone) if someone is going to subpoena police data to prove your history of lateness, leaving work early, etc.
the constitution was not written when police could leave little electronic watchmen to follow your every move - it was intended for a time when actually following someone or surveillance was an expensive proposition requiring knowledge beforehand and a full-time body. We should not turn into a police state just because it's a lot simpler to keep extensive records of every trivial detail. hat's not the spirit of the 4th amendment.
[ link to this | view in thread ]
Re: Re: Re: Re: Re: Re:
Is that really true overall? It seems the complete reverse for some people. Speaking for myself, my 'facial recognition' brain-database is probably much larger, as well as more accurate. Though I'm sure if I became blind I'd develop a higher degree of auditory skills, and hopefully not offend as many people by failing to recognize their telephone voices.
[ link to this | view in thread ]
Re: Re: Re: Re: Re: Re: Re:
You may be highly unusual, but pretty much every study into this topic says that you're mispercieving your abilities.
However, this is a tricky topic. The type of error differs between the two, but generally yes, you're better with voices than faces. When I'm talking about "voice recognition" I mean identifying the identity of a speaker from voice, not understand what the speaker is saying.
We are very, very good at this and can identify the voices of people we know even in very noisy environments or from poor quality recordings with a reasonable degree of accuracy.
The accuracy of identifying people from their faces in similarly suboptimal conditions is much lower. In part, that's because there's a lot less information to go on -- faces have a lot fewer unique identifying characteristics than voices do.
Also, in part, this is because of the difference between identifying things from a static data set versus a dynamic data set. Identifying by voice is dynamic -- we use the rhythm of speech over time as part of the characteristics we're using for identification. A photo of a face is static and as such, we are much more susceptible to false positives: thinking a picture is of someone we know when it's just someone who looks a lot like someone we know. If we're using a movie of their faces as they interact with their environment, our recognition rate improves dramatically.
Interestingly, we are also much, much better at recognizing people based on how they walk than by looking at their photos, even if we're watching them walk from behind them and can't see their faces at all.
In the end, it boils down to this: our brains are the most incredible pattern matching machine we know of. No computer* we've built yet even begins to come close. And the way our pattern-matching works, the more information we have to match with, the better we are at matching (computers* are the opposite of this -- more information makes the task more difficult, not less). Static faces have a lot less information to work with, and results in worse matching by people.
[ link to this | view in thread ]
Re: Re: Re: Re: Re: Re: Re: Re:
*By "computer", I mean von Neumann machines. Other specialized types of computers such as massively parallel or certain analog machines are a whole lot better at pattern matching.
[ link to this | view in thread ]
Re: Re: Re: Re: Re: Re: Re: Re: Re:
There's been research on using a person's ears for identification, since much like fingerprints, it's physical characteristics can be mapped and identified by computers fairly easily.
However the technology evolves, I'm sure the time will come when a stranger walking down the street can be quickly identified by authorities by some kind of scanning-recognition system. Just like license plates.
[ link to this | view in thread ]