EPIC Asks The DOJ To Take A Closer Look At ShotSpotter Deployment
from the misuse-of-federal-funds? dept
ShotSpotter may consider itself to be a solid contributor to the fight against crime, but the facts don’t really bear that out. What it is capable of doing is sending cops to any place a gunshot is (possibly) detected, but that hasn’t really shown to have any meaningful impact on solving gun-related crimes, much less the prevention of future gun crime — the latter of which is far more useful than scrambling cops to some place something may have happened.
ShotSpotter’s brand has taken several damaging hits in recent years. In addition to being dumped by high-profile law enforcement agencies, it has also been accused of altering gunshot records to align more closely with police narratives or justify whatever actions were taken following the initial ShotSpotter report.
Rather than deal with the source of the wounds (all the stuff listed above), ShotSpotter has chosen to re-brand as “SoundThinking,” apparently hoping to fool inattentive Googlers into believing it’s not the same shot-spotting company that has underperformed on a pretty regular basis. It has also decided to dilute its brand by swallowing up the creator of the (also rightfully maligned) predictive policing program, PredPol. Geolitica, PredPol’s creator, is apparently due to be partially consumed by SoundThinking to create a one-stop shop for subpar cop tech.
It had better act fast. ShotSpotter is about to receive another wave of unwelcome scrutiny. As Dell Cameron and Dhruv Mehrotra report for Wired, EPIC (Electronic Privacy Information Center) is asking the DOJ to take a closer look at ShotSpotter deployments.
Attorneys for the nonprofit Electronic Privacy Information Center—a leading US-based civil liberties group—argue that “substantial evidence” suggests American cities are disproportionately deploying an acoustic tool known as ShotSpotter in majority-minority neighborhoods. Citing past studies, EPIC alleges that data derived from these sensors has encouraged some police departments to spend more and more time patrolling areas where the fewest number of white residents live—an allegation disputed by SoundThinking, the system’s manufacturer.
The letter [PDF] sent to Attorney General Merrick Garland cites this data. It also suggests some ShotSpotter deployments are occurring in violation of federal law. The Civil Rights Act — specifically Title VI — is in play here. That section of the law says recipients of federal funds can’t engage in racial discrimination. But that’s exactly what appears to be happening when law enforcement use these funds to purchase and deploy ShotSpotter tech.
First off, EPIC points out that, in addition to any potential Civil Rights Act violations, the tech may just be a waste of taxpayers’ money.
In 2021, the Inspector General of Chicago conducted the most revealing and public evaluation of ShotSpotter’s effectiveness, painting a damning picture of the tool’s impact in light of its significant cost—$33 million for only three years of service. Over 17 months in that period, the Chicago IG reported that over 50,176 ShotSpotter alerts were “confirmed as probable gunshots by ShotSpotter,” each of which resulted in a police response to the alerted area. Only 9.1% of those led to police finding any evidence related to a gun-related crime.
Publicly available data from other cities has only furthered questions about ShotSpotter’s effectiveness. In Houston, more than 80% of ShotSpotter’s 6,300 reports between December 2020 and March 2023 were cancelled, marked as unfounded, dismissed, or closed because officers did not find evidence upon arrival. In Dayton, only 118 of the 2,215 ShotSpotter alerts sent between December 11, 2020, and June 30, 2021, resulted in police reporting an incident with any crime…
That’s what just isn’t working. Here’s why the DOJ might want to take a closer look at this particular use/waste of federal money:
Even relying on ShotSpotter dispatches as a proxy for sensor location, it appears that in practice, ShotSpotter sensors are placed almost exclusively in communities of color. The maps… demonstrate that in Chicago, ShotSpotter sensors and “hits” of “probable” gunshots are heavily concentrated in Black neighborhoods, whereas neighborhoods with more White people were either “not confirmed to have ShotSpotter sensors” at all or had 0-1 sensors.
It’s not just Chicago. EPIC’s study of what available data there is (most ShotSpotter locations are kept secret by the company and its law enforcement customers) shows that there has never been a ShotSpotter alert in Washington, DC’s predominantly white neighborhoods. As the report notes, that’s not because there has been a lack of gunshots or gun crime in these areas. The only conclusion that can be drawn is that the DC PD decided white neighborhoods didn’t need to be dirtied up with this crime-fighting equipment.
Neither of these cities is an anomaly.
In Houston, people of color made up 95% and 80% of the population in the two ShotSpotter pilot program zones.
In Dayton, Ohio, ShotSpotter’s first deployment was in west Dayton, “the heart of the city’s Black community.”
In Kansas City, Missouri, ShotSpotter was deployed in a 3.5-square-mile area of the city in neighborhoods where White people make up only 3.5% of the population.
Newark expanded its use of ShotSpotter into schools in 2021, and 30 of the 34 school buildings selected are in majority Black zip codes in schools with nearly 70% Black students.
ShotSpotter (and its customers) often justify this seeming racial disparity by saying they’re just using data to determine mic placement. But that’s its own problem: the data fed to predictive policing programs and other algorithms to supposedly increase law enforcement “efficiency” is tainted by the years of biased policing that generated this data. What’s treated like math and science is usually just a self-fulfilling prophecy: if cops believe certain areas (or certain people) are more prone to crime, they will spend more time there, generating even more data to feed to machines that pump out little more than byte after byte of confirmation bias.
The letter asks the DOJ to take a close look at funding recipients and their use of this tech. It also suggests the federal government would be well-served to suspend any further funding of surveillance tech until it has this sorted out and has put meaningful limits, guidelines, and other tools of accountability in place to prevent future misuse of federal funds.
The DOJ has yet to respond or comment on EPIC’s letter or its suggestion some federal fund recipients are violating Title VI requirements with biased deployments. But it has prompted one of the most dogged critics of all things surveillance to respond with his personal assurance that both ShotSpotter and the DOJ won’t be allowed to simply pretend this has never happened.
Ron Wyden, a leading privacy hawk in the US Senate, tells WIRED that he planned to urge Garland to adopt EPIC’s recommendations.
“There is more than enough evidence at this point to conclude that technologies like ShotSpotter do essentially nothing to stop crime,” Wyden says, “but instead have a well-documented discriminatory impact on marginalized and vulnerable communities.”
And if that’s all ShotSpotter really is: an expensive add-on for the same old racist policing, one has to wonder why the DOJ is bothering to fund it at all. If nothing else, this might make SoundThinking a little less popular with local PDs, if only because they can’t get the federal government to buy these useless toys for them.
Filed Under: civil rights, doj, privacy
Companies: epic, shotspotter
Comments on “EPIC Asks The DOJ To Take A Closer Look At ShotSpotter Deployment”
But see, white people never commit crimes if you never look.
I’m always reminded of how people used to tell me that more non-whites got arrested for weed use because whites didn’t do it… All while I knew a ton of white people who would bring weed into their college classes…
Re: maybe check the stats?
https://www.ncbi.nlm.nih.gov/books/NBK568206/figure/ch5.fig6/
Re: Re: Maybe look at the correct stats?
I don’t know why you chose that chart, but it does fit in with using select stats to prove your point to get select stats.
First look at your link does seem to backup your point. There are two things I noticed though: 1) only 1.3% of white and 2.4% of black populations use weed? 2) the title uses the word “disorder”.
After hunting around in the site I found that “disorder” is not usage, it is defined to have problem with usage including encounters with law enforcement.
Then looking for the usage chart (same chapter just a little above): [https://www.ncbi.nlm.nih.gov/books/NBK568206/figure/ch5.fig3](https://www.ncbi.nlm.nih.gov/books/NBK568206/figure/ch5.fig3) you can see usage without the law enforcement and the numbers are more like 15% and 17% for each which I would believe.
The above shows that usage is close but “disorder” with law enforcement tends to be double. Using the “disorder” as proof law enforcement should be doing more is a circle of logic.
This comment has been flagged by the community. Click here to show it.
Well, if black men would stop shooting each other with such frequency, there’d be no need to focus on their neighborhoods.
Re:
hi hyman
Re:
Didn’t read the posting, or just ignoring what it had to say about your “point”.
Was wondering why Epic was asking the DOJ about gunshot tech… Turns out its the other EPIC and Epic will stick to games.
“ShotSpotter sensors are placed almost exclusively in communities of color.”
Have you noticed that perhaps they are placed “almost exclusively in communities” with high homicide rates, which just so happen to be considerably higher in “communities of color”.
This is the same tired trope, of blaming people who want to stop crime, when the victims of those crimes are mostly “communities of color”. If the sensors aren’t working, then design better sensors, but don’t blame people for trying to fix violent crime, through non invasive means.
If anything it just means they should erect more sensors, and design a better machine learning model, so that they can better filter out the false positives, and maybe cite more people for not maintaining their mufflers.
Re:
Tunnel vision is very limiting, it’s like you can not see other stuff right there in front of you.
Must be very difficult for you.
Re:
Hahahah.. cops.. are trying to solve violent crime?
Please read ANY info on shotspotter. You are completely ignorant of it and look like an idiot. False positives aren’t an issue, manually edited data sometimes after the fact so that cops can get away with their murder is.
There is a pretty easy way to get rid of violent gun crime, get rid of guns. Every single other solution is just about catching the person after the fact, or keeping the body count down.
Re: Re:
How do you get the guns out of the criminals hands? They tend to not give a crap about gun bans or following laws.
How will law abiding citizens protect themselves from gun toting criminals if all guns are banned? Police are always minutes away and taking a knife to a gun fight usually does not end well for the person with the knife.
Guns exist, they are simple to manufacture, and the knowledge of how they work is well known. You cannot ban them and make them all disappear, if someone wants one bad enough they will make one or buy one from someone who can make them.
Perhaps we can make all machine lathes, mills and 3D printers illegal too so no criminals can make their own guns.
We could also ban all materials used to make the above noted tools so criminals cannot make them to make guns.
Better ban books and dissemination of information too since criminals might share instructions on how to make a gun.
Other deadly weapons that are easier to make than guns exist too, criminals could just use those instead and some of them are already banned yet used by criminals.
Re: Re: Re:
Outlaw doing something, and only the outlaws will do it.
Re:
Detecting gunshots and sending in the cops like they are the Calvary will do very little to reduce crime, but rather make the people living in those districts feel even more oppressed and cut off from the rest of society. Reducing crime will take a real effort to reduce inequalities and improve the economic opportunities for people.
I look at it as technology that is still in infancy status. With the labor shortages that we have, increasing productivity from the police is imperative, and technology is the only way that is going to happen.
So, we don’t need to get rid of shot-spotter, we need a better version. Same with facial recognition.
Re:
i would rather just have better cops. So far, “labor shortages of cops” don’t seem to be a bad thing.
P.S. The better version of shotspotter would be an installation of a regular grid of detectors covering everywhere. (Better software, detectors, and operators notwithstanding.)
That’s because you’re probably a Hamas-supporting communist.
Data manipulation
This alone should disqualify the whole program until an audit, and probably major changes, have been conducted.
When you add the discriminatory nature of sensor location, this whole thing is not just a waste of money, it perpetuates a system of minority oppression by law enforcement. It’s literally worse than nothing.
ShotSpotter
I think there own service agreement says it all and I quote ” “ShotSpotter does not warrant or represent, expressly or implicitly, that the Software or Subscription Services or its use will: result in the prevention of crime, apprehension or conviction of any perpetrator of any crime, or detection of any criminal; prevent any loss, death, injury, or damage to property due to the discharge of a firearm or other weapon”
ShotSpotter is paying $ to racially profile communities and cause harm.