I was imagining selective enforcement/response, or restricting use to areas where one race predominates, or behaving differently in response to mic reports in white neighborhoods than black. Standard cop shit.

edit: yeah pretty much

Well they are placed in mostly minority communities but above that, a ShotSpotter tech admitted in court that they often change the analyst interpretations of what is and isn't a gunshot at the request of their police department customers which, keep in mind, have successfully been used in court as evidence of a crime.

There is also overwhelming data showing that the majority of their alerts lead to no arrests. The Chicago IG believes this demonstrates false positives where as ShotSpotter (who changed their name after some criticism to SoundThinking) says that people can fire a gun and leave no evidence in spite of police investigating and asking people that would have witnessed it (and there being no victim).

Furthermore, the way it works is that AI 'assists' a human who then determines if it is a gun shot and then attempts to triangulate the position it came from. From the trial I mentioned earlier, ShotSpotter determined that there was no gun shot but then changed the analysis at LEO request but in actuality was proven to be a helicopter....

Furthermore, ShotSpotter keeps the details of its methodologies and models a secret and has refused an independent audit from IPVM.

So with all of that, one could easily argue that ShotSpotter/SoundThinking is as biased as a police officer and that the evidence is purely subjective and non-transparent.

By putting them only in predominantly minority neighborhoods. And I do mean only.

The bias comes in because they installed them in areas with more gunshots, which happen to be areas with more minorities.

By not spending excessively on detectors uniformly distributed across the metropolitan area, they are targeting minorities.