ShotSpotter Looking To Compound Bad Cop Tech Ideas By Acquiring Predictive Policing Software Company
from the unholiest-of-alliances dept
ShotSpotter has long presented itself as a reliable detector of gunshots. Mileage, however, has varied. Law enforcement customers that have gotten disgruntled with this limited service have pointed out — en route to terminated contracts — that (a) detected gunshots are not always gunshots, (b) detected gunshots are rarely useful intel, and (c) detecting gunshots rarely results in less gun-related crime.
The law enforcement agencies that still buy into ShotSpotter like it for other reasons. They like the fact that minimally-trained “analysts” will alter reports to provide cops with post facto justification for their actions. They like the “reasonable suspicion” ShotSpotter’s shot detection supposedly provides — justification for stopping and searching anyone in the area of a reported shot. And, presumably, they like ShotSpotter’s willingness to sue journalists who paint a negative picture of the company by reporting on court documents, testimony from its own experts, and the company’s internal documents.
ShotSpotter has dragged its own name through the mud in recent years, resulting in a recent rebranding as “SoundThinking,” apparently hoping Googlers will look past search results that contain phrases like, “SoundThinking, formerly ShotSpotter.” That’s the sort of SEO-washing companies do when their original names become a bit too toxic, like Taser (now Axon) or, say, military security contractor Blackwater, which rebranded to Xe Services (2009), Academi (2011), and finally Constellis after a merger with another private security firm.
ShotSpotter (a.k.a. SoundThinking) is now hoping to add to its questionable tech arsenal by acquiring yet another rebranded firm after it found its original name inextricably tied to a bad product with an extremely tainted reputation (at least outside of the law enforcement world). Here’s Dell Cameron and Dhruv Mehrotra reporting on the sort of corporate marriage that should have plenty of people shouting their objections when asked by the person presiding over the ceremony.
SoundThinking, the company behind the gunshot-detection system ShotSpotter, is quietly acquiring staff, patents, and customers of the firm that created the notorious predictive policing software PredPol, WIRED has learned.
In an August earnings call, SoundThinking CEO Ralph Clark announced to investors that the company was negotiating an agreement to acquire parts of Geolitica—formerly called PredPol—and transition its customers to SoundThinking’s own “patrol management” solution.
“We have already hired their engineering team,” Clark said during the call, a transcript of which is public. He added that the acquisition of patents and staff would “facilitate our application of AI and machine learning technology to public safety.”
Admittedly, PredPol is a terrible name for something used by law enforcement. But it’s the concept that sucks the most: the presumption that feeding AI a constant onslaught of tainted data will somehow result in better policing, rather than the racist garbage version we’ve endured for decades. The bigoted past will somehow become a level playing field… because computers. That’s not science. That’s not even faith. That’s straight-up delusion.
No wonder PredPol rebranded as Geolitica. It had to. But it possibly rebranded too early, which means it may not be able to outrun its history of failure.
In December 2021, Gizmodo and The Markup analyzed millions of Geolitica’s crime predictions that were discovered on an unsecured server and found that the software disproportionately—and often relentlessly—targeted low-income communities of color for additional patrols.
And, like ShotSpotter’s reputation-making tech, predictive policing software has been rejected by major law enforcement agencies in recent years after those using it discovered it changed almost nothing about the way resources were deployed. If you’re just going to cop the way you’ve always copped, there’s no reason to be shelling out millions of dollars a year for artificial intelligence capable of little else but confirmation bias.
This acquisition would put all the bad eggs in one basket, or — to paraphrase a colloquialism cop officials seem to believe is exonerative — all the bad apples in one barrel.
Now, there are obviously reasons to be very concerned about this acquisition. Putting all this under one roof means both companies will be able to absorb contract losses and critical press more easily. It also means both products are likely to get worse, since there will be little motivation to improve offerings and/or create checks and balances to offset the incoming flow of biased policing data.
On the other hand, it also means that when data is exposed — like the data analyzed by Gizmodo and The Markup — there should be much more of it. That’s going to be extremely useful to researchers and activists who have long recognized the downsides of letting software do cops’ jobs for them. And this merger will create an extremely enticing target for transparency enthusiasts who hunt down exposed data and/or are willing to phish someone into handing over the keys to the database.
Not exactly win-win. But better than a straight-up loss for the general public. To be sure, the people already routinely oppressed by law enforcement will likely see no change in the level of oppression… at least not immediately. But even this consolidation of artificial police “intelligence” contains a silver lining: when it’s exposed for the artifice it is, it will be spectacular.