Gunshot detection technology known as ShotSpotter has been used as evidence in 190 court cases even as some critics raise questions about the technology.
According to Vice’s Motherboard, prosecutors are using the ShotSpotter results despite a lack of empirical evidence about the accuracy of its sensors or its impact on crime deterrence. Another problem is that ShotSpotter analysts sometimes modify results after police departments request a review.
ShotSpotter is being used in more than 100 cities, including Chicago and New York City.
The article points to a Chicago case in which prosecutors withdrew ShotSpotter evidence after public defenders claimed in a motion that ShotSpotter analysts overrode the results.
According to the assistant Cook County public defenders, Lisa Boughton and Brendan Max, a ShotSpotter analyst reclassified a sound deemed to be fireworks to conclude that it was a gunshot. Another analyst changed the location coordinates from 5700 South Lake Shore Drive to a location a mile away on South Stony Island Drive, their motion says.
Prosecutors had relied on the evidence to charge 64-year-old Michael Williams with the shooting of a 25-year-old man he dropped off at a hospital. Police had contended that Williams’ car was recorded on a camera at the Stony Island Drive location at the time of the sound picked up by ShotSpotter.
“Through this human-involved method, the ShotSpotter output in this case was dramatically transformed from data that did not support criminal charges of any kind to data that now forms the centerpiece of the prosecution’s murder case against Mr. Williams,” the public defenders wrote.
If the evidence had not been withdrawn, Illinois courts would have examined ShotSpotter science and source code for the first time, according to Jonathan Manes, an attorney at the MacArthur Justice Center.
“Right now, nobody outside of ShotSpotter has ever been able to look under the hood and audit this technology,” Manes told Vice’s Motherboard.
ShotSpotter has claimed accuracy ranging from 80% to 97%. ShotSpotters’ own employee, Paul Greene, told a San Francisco court in 2017 that its accuracy guarantee “was put together by our sales and marketing department, not our engineers.”
Greene also acknowledged in two other cases that he had reevaluated ShotSpotter results after requests from police departments, according to court documents cited by Vice’s Motherboard.
Sam Klepper, senior vice president for marketing and product strategy at ShotSpotter, told Vice’s Motherboard that ShotSpotter analysts “agree with the machine classification over 90% of the time.” But in “a tiny number of cases, our customers request us to perform a location analysis to validate the accuracy of the location. If we find an error, we provide a more accurate location to the customer to assist the investigation.”
Klepper also told Vice’s Motherboard that Greene’s testimony about accuracy “had nothing to do with the determination of our actual historical accuracy rate. While marketing and sales have appropriate input on our service level guarantees for our contracts, actual accuracy rates are based on detections that we record.”