Privacy advocates respond to leaked data of ShotSpotter gunshot detection

Privacy advocates say a leak of data revealing the locations of ShotSpotter sensors — used by law enforcement to detect and locate gunshots — should support arguments for cities to reconsider their use of such technologies.

Information from the leaked data appeared in a report by WIRED, exposing for the first time the locations of individual ShotSpotter sensors, which were always closely guarded secrets by both city governments and the ShotSpotter company, now called SoundThinking.

Privacy advocates are continuing to process the leak, which has taken aback even activist groups who’ve organized around ShotSpotter for years. The leak, reported  in WIRED last week, revealed that over 12 million Americans live in neighborhoods with at least one ShotSpotter sensor. Those sensors can be found in 84 metropolitan area, across 34 states and U.S. territories. Nearly 70% of people who live in those neighborhoods are Black or Latino, the data showed, with nearly three-quarters of those neighborhoods being majority nonwhite, and with average household annual incomes of around $50,000 a year.

The picture wasn’t much different in Milwaukee, which has used and grown a network of ShotSpotter sensors since 2010. In 2022, Wisconsin Examiner reported that since 2010, the city of Milwaukee has spent $3.7 million on ShotSpotter. By 2023, Urban Milwaukee reported, the city and ShotSpotter had signed $5.25 million contract lasting until March 2026. Portions of the contract were offset by state grants, the news outlet noted.

ShotSpotter works by using audio sensors deployed in neighborhoods to detect loud, impulsive sounds. Those could include gunshots, fireworks or  cars backfiring. Representatives from ShotSpotter told Wisconsin Examiner in 2022 that the system digitally disregards sounds that are likely not gunshots, such as fireworks or helicopters. The remaining sounds are then sent to a ShotSpotter technician, who determines whether the sound clip is of gunshots. Afterwards, the technician either sends an alert to local police with a triangulated location of the gunshot or debunks the alert as not being from a gunshot.

All that occurs within 60 seconds after a ShotSpotter sensor is activated. In Milwaukee, the system is overseen by the Milwaukee Police Department’s intelligence fusion center, which includes the homeland security-focused Southeastern Threat Analysis Center. Once a ShotSpotter sensor detects a gunshot, the fusion center can activate nearby cameras the MPD has access to before  a squad car arrives on the scene. The police also may deploy drones to surveil an area after a sensor is activated. Over 14,000 activations of the ShotSpotter system occurred last year in Milwaukee.

MPD is the only city police department in Wisconsin using ShotSpotter sensors. Prior to the leak, MPD would only identify which police districts used ShotSpotter sensors. Those districts encompassed predominantly African American portions of the North Side, and largely Hispanic and Latino neighborhoods on the South Side. In open records denials, the department claimed that revealing the exact location of ShotSpotter sensors would hinder MPD’s ability to respond to crime and endanger private citizens who’ve agreed to host ShotSpotter sensors on their own property.

MPD wouldn’t comment on the leak and directed all questions to SoundThinking, which the company has renamed itself.  In a written statement, a SoundThinking spokesperson said the leak included “high-level ShotSpotter sensor locations for several jurisdictions” and that reporting by WIRED “has been the basis for several additional stories on this matter — including yours.”

The spokesperson, who asked not to be named, added: “We want to be clear that SoundThinking believes the document containing this confidential information was illegally disclosed by ex-employees and is currently pursuing civil and criminal remedies against the private parties responsible. Due to this ongoing litigation, we cannot comment specifically on the leaked data, however, we will continue to object to the use of our stolen data and reinforce the dangers of disclosing individual sensor locations.”

The spokesperson said public knowledge of the sensor locations “represent a security and operation risk — towards this end, based on our guidance — WIRED did acknowledge the risk and agreed not to disclose individual sensor locations.”

Questions beyond the leak

“I think it’s kind of dystopian that there are over 25,000 microphones in communities nationwide,” Jon McCray Jones, policy analyst at the American Civil Liberties Union (ACLU) of Wisconsin, told the Wisconsin Examiner.  Jones, who specializes in LGBTQ+, surveillance and criminal justice reform policy, was taken aback by the scale of ShotSpotter programs nationwide. A program on that scale, while chilling to some, also follows “America’s long history in over-surveilling Black people, brown people and people from low income neighborhoods,” said Jones. “You’re over-surveilling the most heavily marginalized communities in the country.”

Whether the system is effective has been a matter of dispute in Milwaukee. Urban Milwaukee reported recently that neighborhoods with the highest number of homicides and non-fatal shootings were the neighborhoods with ShotSpotters, citing the MPD’s crime map. However, the outlet also noted that portions of the northwest corner of the city do not have ShotSpotter sensors, despite also having substantial shooting rates.

MPD’s online crime data only goes back to 2022. By comparison, ShotSpotter has been active in Milwaukee since 2010 through multiple peaks and valleys of in the rate of homicides and gun violence. Homicides in Milwaukee peaked in 2015, following the loosening of gun laws, then gradually decreased until 2020. In early 2022, TMJ4 reported that ShotSpotter had been deployed largely in Black neighborhoods, despite other more diverse districts having more gunshots. That same year, when asked by Wisconsin Examiner, neither the MPD nor ShotSpotter representatives could provide data showing the system directly contributed to improving gun violence, or solving homicide cases, in Milwaukee.

Jones said the data published by WIRED mirrors the reality on the ground in Milwaukee. “The communities being surveilled by ShotSpotters are Black and brown communities,” said Jones, who pushed back against assumptions that those are simply the neighborhoods with the most gun violence. “From the map, we can’t tell what streets these ShotSpotter microphones are at. So I think that it might be inaccurate, if not irresponsible, to say that these technologies are present in the spots that experience the most gun violence, because we just don’t have the precise location …  of these microphones to know for sure that that’s where they’re located.”  Saying that the technologies are where they are supposed to be to keep communities safe, he added is  “more pushing a narrative.”

MPD confirmed to  Urban Milwaukee last week that the department chooses where ShotSpotters are deployed. SoundThinking told Wisconsin Examiner that the sensors “are deployed based on objective criminal gunfire data and, despite the claims that ShotSpotter alerts lead to bias and over-policing in predominantly Black and Brown communities, ShotSpotter provides intelligence that allows police to coordinate safe, efficient and equitable responses that require fewer resources in a way that builds community trust.”

SoundThinking’s spokesperson contended that by detecting gunfire, ShotSpotter helped ensure that law enforcement and first responders “know about it and can respond quickly to render aid to victims and secure the area.” ShotSpotter alerts give police “more information and context about the incident than they would have without ShotSpotter,” the spokesperson said. “This information allows for a safer and more equitable response.”

Jones, and other privacy advocates disagree with that premise. “Taxpayers in Milwaukee are paying millions of dollars to the ShotSpotter company, and we don’t know if this technology makes us safer,” Jones told Wisconsin Examiner. “If we’re going to trade our privacy to law enforcement and to the state, the community should know that the privacy that they’re trading comes with some type of benefits. And the fact that ShotSpotters in the city of Milwaukee can’t produce any evidence that this privacy that we’re giving up comes with benefits to the city or benefits to the communities, and the fact that most of these taxpayers don’t even know about the technologies being imposed on them, [are] extremely problematic.”

Studies on ShotSpotter have provided mixed results. In the past, studies touted by the technology’s manufacturer suggested that ShotSpotter increases police awareness of gunshots, streamlined emergency responses, and that in some areas assaults and gun crime dropped after the system was deployed.

On the other hand, a report by the Macarthur Justice Center found that 89% of ShotSpotter alerts in Chicago did not lead to police finding evidence of gun crime. The report also found that 86% of ShotSpotter alerts led to no report of a crime being filed, and argued that the technology could fuel more confrontational encounters with police in vulnerable communities. Another study  in St. Louis found that ShotSpotter did not deter gun violence, with similar findings in  Philadelphia involving gunshot detection technology that  wasn’t branded ShotSpotter.

Growing demand for oversight of police surveillance

It’s unclear what measure city officials or police will take in response to the leak. Re-locating every sensor, for example, would be a costly and slow process. Campaign Zero, which has organized over the last couple of years to persuade  cities to cancel their contracts with ShotSpotter, said that the leak’s findings should alarm local leaders.

“City officials should remove ShotSpotter completely since the technology has been proven to harm communities by manufacturing probable cause, leading to false arrests and wrongful incarceration, without reducing gun violence,” the group wrote in a statement to Wisconsin Examiner. “While we encourage transparency and the release of ShotSpotter sensor locations so that independent researchers can analyze the company’s claims, cities would be better served divesting from fault technology that does not improve public safety and investing in communities.”

Jones argues that the city of Milwaukee should adopt policies such as  Community Control Over Police Surveillance (CCOPS) ordinances. “Basically what CCOPS does is, every police surveillance technology that is imposed on the public gets a public hearing,” explained Jones. Community residents, company representatives, elected officials and the police could offer testimony on whether the technology should be used in the city. The ordinance would also create an annual report detailing spending on surveillance technologies and statistics on how the technology has improved public safety. “Each year, it gives the community the ability to opt out of these technologies that we’re frankly footing the bill for,” said Jones. “I think that’s how we need to handle surveillance in general.”

Such an ordinance could also provide an alternative path for oversight after city officials surrendered Milwaukee’s Fire and Police Commission’s ability to set policy and provide oversight for the police department in exchange for a new sales tax. A CCOPS ordinance would essentially position the city’s common council “to become the new Fire and Police Commission,” said Jones, with the council’s Public Safety Committee holding hearings. “The community gets to voice their opinion, the alder people don’t get to say ‘Hey we didn’t know this was happening, we didn’t know about these technologies,’” said Jones. “Now all of our hats are in the game for the surveillance. It is not just something that is happening to communities without the knowledge of all parties.”